Skip to main content

Full text of "Bulletin of the Museum of Comparative Zoology at Harvard College"

See other formats


special Collections 

H05 



IjM^f) 






^'> not circulate 



HARVARD UNIVERSITY. 




LIBRARY 



OF THE 



MUSEUM OF COMPARATIVE ZOOLOGY 



iBBB' 







NOV 8 1933 



BULLETIN 



OF THE 



MUSEUM OF COMPARATIVE ZOOLOGY 



AT 



HARVARD COLLEGE, IN CAMBRIDGE 



VOL. LXXIV 



CAMBRIDGE, MASS., U. S. A. 
1933 



-^ / " 



The Cosmos Press, Inc. 
Cambridge, Mass., U. S. A. 



ft-, i 



CONTENTS 



PAGE 



j^Q 1 — On the Blood Vascular Bundles in the Limbs of Certain 
Edentates and Lemurs. By George B. WLslocki and 
William L. Straus, Jr. (4 plates). April, 1932 ... 1 

No. 2. — A Study of the Osteology of Alligator Prenasalis 

(Loomis). By Charles C. Mook. (3 plates). July, 1932 17 

No. 3. — On Certain Similarities between Sloths and Slow 
Lemurs. By William L. Straus, Jr. and George B. Wislocki. 
September, 1932 43 

No. 4. — Fossil Types of Fishes, Amphibians, Reptiles and Birds 
IN the Museum of Comparative Zoology. By W. E. 
Schevill. December, 1932 57 

No. 5. — Birds from Northwest Yunnan. By James C. Greenway, 

Jr. February, 1933 107 

No. 6. — New and Little Known Spiders from the United States. 

ByElizabethB. Bryant. (4 plates). June, 1933 . 169 

No. 7. — Reports on the Scientific Results of an Expedition to 
the Southwestern Highlands of Tanganyika Territory. 
VIL Herpetology. By Arthur Loveridge. (3 plates). 
October, 1933 195 -^/'^ 



T-.-YI .^^-^ ^j \l<ji^^^^ 'Jot, 7^r. 



APR 1 8 1932 



Bulletin of the Museum of Comparative Zoology 

AT HARVARD COLLEGE 

Vol. LXXIV, No. 1 



OK THE BLOOD VASCULAR BUNDLES IN THE LIMBS 
OF CERTAIN EDENTATES AND LEMURS 



By George B. Wislocki and William L. Straus, Jr. 



[From the Departments of Anatomy, Harvard University and 
Johns Hopkins University] 



With Four Plates 



CAMBRIDGE, MASS., U. S. A.: 
PRINTED FOR THE MUSEUM 
April, 1932 



PUBLICATIONS 
OF THE 

MUSEUM OF COMPARATIVE ZOOLOGY 
AT HARVARD COLLEGE 



There have been published of the Bulletin Vols. I to LXV, LXVII- 
LXXIII; of the Memoies Vols. I to LI. 

The Bulletin and Memoirs are devoted to the publication of 
original work by the Officers of the Museum, of investigations carried 
on by students and others in the different Laboratories of Natural 
History, and of work by specialists based upon the Museum Collec- 
tions and Exploration. 

These publications are issued in numbers at irregular intervals. 
Each number of the Bulletin and of the Memoirs may be sold sepa- 
rately. A price list of the publications of the Museum will be sent on 
application to the Director of .the Museum of Comparative Zoology, 
Cambridge, Massachusetts. 



Bulletin of the Museum of Comparative Zoology 

AT HARVARD COLLEGE 

Vol. LXXIV, No. 1 



ON THE BLOOD VASCULAR BUNDLES IN THE LIMBS 
OF CERTAIN EDENTATES AND LEMURS 



By George B. Wislocki and William L. Straus, Jr. 



[From the Departments of Anatomy, Harvard University and 
Johns Hopkins University] 



With Four Plates 



CAMBRIDGE, MASS., U. S. A.: 
PRINTED FOR THE MUSEUM 
April, 1932 



No. 1. — On ihr Blood VascxiJar Bundles in the Limbs of Certain 

Edentates and Lemurs 

By George B. \Yislocki and William L. Straus, Jr. 

[From the Departments of Anatomy, Harvard University and 
Johns Hopkins University] 

Introduction 

At the beginning of the last century (1800, 1804), Carlisle wrote 
two short papers on " a peculiarity in the distribution of the arteries 
sent to the limbs of slow moving animals," in which he described, in 
the extremities of sloths, as well as of "Lemur tardigradus" and of 
"Lemur loris," blood vascular plexuses, related, as he believed, to the 
slow movements executed by these animals. Now, over a century later, 
great uncertainty still exists as to the significance and even the mor- 
phology of the various forms of such plexuses, or retia mirabilia, found 
in various mammals, especially in their extremities. 

Weber (1928) is of the opinion that it is inadmissible to correlate 
the blood vascular plexuses of the extremities of the Bradypodidae 
with their slow movements, because plexuses are present in such active 
animals as the Dasypodidae (2, p. 27). Concerning the lemurs he says 
that in the extremities there occur arterial and venous plexuses similar 
to those seen in the Xenarthra. These plexuses are found in the Lori- 
sinae, which during the day sleep on the branches of the trees, while at 
night they execute their sluggish movements. It would be hazardous, 
however, according to Weber (2, p. 732), to associate the slow move- 
ments of these animals with the plexuses, because Vrolik (1826) has 
described similar plexuses in Tarsius, a nocturnal animal which is 
extremely active. 

We have had the opportunity of studying both grossly and micro- 
scopically the blood vessels of the extremities of a number of speci- 
mens of edentates and of lemurs and of Tarsius. Our observations, we 
believe, help to clarify certain of the problems relating to the retia 
mirabilia in the hmbs of mammals. 

Alaterial 

Of the Xenarthra we have had at our disposal for dissection numer- 
ous specimens of Choloppus hoffmanni, the two-toed sloth, and of Brady- 
pus griseus, the three-toed sloth; one specimen of Cyclopes didactylu^, 



4 bulletin: museum of comparative zoology 

the two-toed ant-eater; one specimen of Tamandua tetradadyla, the 
four-toed ant-eater; and several specimens of Dasypus notemdnctus, 
the nine-banded armadillo. 

Of the Pholidota we have dissected one specimen of Mauis javanica, 
the pangolin. 

Of the Lemuroidea we have studied a single specimen of Perodicticus 
potto, and two specimens each of Nycticebvs borneanus, Lemur varie- 
gatiJS and Galago sp.? 

Of the Tarsioidea we have examined single specimens of Tarsius 
phihppinensis, T . fraterculns and T . saltator. 

We wish to express our thanks to ^Ir. Gerrit S. Miller, Jr. of the 
U. S. National IMuseum and Dr. Milton J. Greenman of the Wistar 
Institute for the use of specimens. We are also indelited to the Barro 
Colorado Island Laboratory, Panama Canal Zone, for the opportunity 
of studying sloths. 

Ohserrntions 

Our observations have added to the evidence that there are several 
types of blood vascular plexuses, as reported by Hyrtl and by Miiller. 
Not only have we examined these specimens by gross dissection, but 
we have also introduced microscopic examination in some instances 
as a more reliable means of discriminating the types of plexuses. 

We have been able to distinguish at least two types of plexiform dis- 
tribution of blood vessels in the extremities: (1) the vascular bundle, 
and (2) the simple vascular network. By vascular bundle we mean a 
type of plexus typically seen in the limbs of sloths. The precise struc- 
ture of such a plexus consists in a breaking up of the main artery of the 
extremity into a principal trunk surrounded by upwards of 30 to 40 
smaller arteries which are gi^'en off from the main stem and accompany 
it as parallel vessels lying in a common sheath. These vessels are 
destined, after running in the sheath for variable distances and an- 
astomosing sparsely, to supply the muscles of the extremity. The veins 
running in the bundle are of about the same number and distribution 
as the arteries. 

The vascular bundle type of plexus occurs in its most pronounced 
form among the Xenarthra, in the didactyl and tridactyl sloths 
{Chohrpvs and Bradypvs) and in the two-toed ant-eater {Cyclopes) 
(figs. 1-4). In these animals the bundle formation involves not only 
the main vessels of arm and thigh, but the chief trunks of forearm and 
leg as well. 

Among our prosimian material we find the same arrangements in 



WISLOCKI AND STRAUS: BLOOD VASCULAR BUNDLES 5 

our examples of the Lorisinae, namely in Nycticrbus and Pcrodicticus. 
In both of these we have observed, grossly as well as microscopically, 
vascular bundles in arm and forearm and thigh and leg; these bundles 
are almost completely identical with those observed in Choloepus, 
Bradiipus and Cyclopes. In the brachial bundle of Pcrodicticus (fig. 5), 
1 large artery, 47 small ones and 28 veins are counted in a typical cross- 
section; in the femoral bundle (fig. 6), 2 large arteries, 55 small ones 
and 39 veins. In Nycficcbus, in the brachial bundle (fig. 7), 1 large 
artery, 27 small ones and 22 veins are present; in the femoral bundle 
(fig. 8), 1 large artery, 59 small ones and 40 veins. The small arteries 
are all of nearly equal size, as are the veins also. It is observed, more- 
over, from the microscopic examination, that anastomoses of the veins 
are more numerous than of the arteries. No single large venous trunk 
corresponding to a brachial or femoral vein occurs in the bundle. The 
femoral bvmdle of Perodicficus in the present specimen, as well as both 
brachial and femoral bundles in Cyclopes, contain two, instead of one, 
main arterial vessels. In the majority of animals possessing bundles, a 
single main artery appears to be the rule. Whether in the present 
instances of Pcrodicticus and Cyclopes the occurrence of two arteries 
is a variation from the norm, or whether this is the usual arrangement 
for these animals, we have no means of ascertaining. 

We also find the bundle type of rete mirahilc occurring in the limbs 
of Tamandua and Manis. In these animals, however, the bundles are 
limited to vessels supplying the forearm and the leg in the respective 
extremities. The great arteries of the upper arm (brachial) and the 
thigh (femoral) are not plexiform as in the sloths, Cyclopes and lorises, 
being single vessels (figs. 13, 14, 16, 17). The reiia of Tamandua and 
Manis prove, however, on both gross and microscopic examination, to 
be true vascular bundles, identical in their structure with those of 
sloth and loris (figs. 15, 18). 

In the other animals examined we found no indications of the 
bundle type of plexus. Dasypus possesses retia mirahilia in its ex- 
tremities, but these are quite dissimilar to the bundles seen in the 
sloths, lorises, ant-eaters and pangolin. They are at most but simple 
networks of the nature of a few widely scattered anastomoses. 

A rather different arrangement obtains in our specimens of Tarsius. 
Sections through the upper arm show that the brachial artery is a 
single tube unaccompanied by a blood vascular bundle (fig. 11). The 
femoral artery, however, instead of being represented by a single vessel, 
breaks up into a number of branches immediately after emerging from 
the pelvis onto the thigh (fig. 12). The plexus so formed is, however, 



6 bulletin: museum of comparative zoology 

by no means like the vascular bundle encountered, for example, in 
sloths. It consists, instead, of arteries and veins, all of about equal 
size, irregularly scattered in small groups throughout the connective 
tissue without the formation of a common sheath. In a typical cross- 
section through the region of the upper part of the thigh are counted 
24 arteries and 28 veins which are scattered over a considerable area. 
Thus in the thigh of Tarshis we are dealing not with a characteristic 
vascular bundle, but with a more diffuse type of plexiform distribu- 
tion. 

In specimens of Galago and Levmr variegatus also, we find, both 
grossly and microscopically, no close resemblance of the vessels of the 
extremities to those of sloths, lorises or two-toed ant-eaters. The blood 
vessels are, in so far as they are plexiform, of the simple network pat- 
tern, and nothing comparable to a vascular bundle involving either 
the upper or lower extremity arteries can be seen. (Figs. 9, 10.) 

By the reinvestigations of these several animals we believe that we 
can safely speak of the blood vascular bundle involving the entire limb 
(arm and forearm, thigh and leg) as characterizing the sloths, the two- 
toed ant-eater and the Lorisinae. A less complete form of this vascular 
bundle, involving only branches to the forearm and thigh, occurs in 
the four-toed ant-eater and pangolin. Simple plexiform anastomoses 
occur in both extremities of the armadillos and in the thigh of Tar sins. 
In Galago and Lemur the main blood vessels of both limbs are in general 
single tubes. 

Discussion 

Plexuses of blood vessels in the limbs of mammals have aroused the 
interest of numerous observers since the time of Carlisle (1800, 1804). 
Thus we find descriptions of them in the papers of Vrolik (1826), 
von Baer (1835), Burmeister (1846), Rapp (1852), Hyrtl (1853, 1864), 
Huxley (1864), Chapman (1874), Zuckerkandl (1895) and Miiller 
(1905). Retia mirahilia of variable degrees of complexity are described 
in representatives of nearly all groups of mammals, for instance in the 
monotremes, some marsupials, all edentates, some ungulates, hyra- 
coids, certain fissipeds, pinnipeds, some rodents, lemurs, Tarsius, and 
finally in the rudimentary arms of cetaceans. 

Both HNTtl (1853, 1864) and Miiller (1905) dissected large compara- 
tive series of mammals. Hyrtl divided retia mirahilia, as he observed 
them in the extremities, into two categories: (1) radially diffuse net- 
works, and (2) massive networks. A similar classification was adopted 
by Miiller, who recognized (1) the network, and (2) the radiating fan 



WISLOCKI AND STRAUS: BLOOD VASCULAR BUNDLES / 

or brush. These investigators thus agree that there are two types of 
rctia mirabilia in the Umbs of mammals. With this our own observa- 
tions accord most fully. We have preferred to speak of these types of 
retia as (1) the simple vascular network, and (2) the vascular bundle. 
A third type, a radiating plexus of blood vessels, as encountered in the 
upper extremities of cetaceans and pinnipeds, deserves perhaps to be 
set apart from the others, but the anatomical data concerning it do not 
justify its consideration at the present time. 

In spite of the fact that both Hyrtl and Miiller define two types of 
retia mirabilia and give illustrations of each, other comparative anato- 
mists have usually ignored their classifications, speaking merely of 
blood vascular plexuses in the extremities of various mammals with- 
out making it clear to which type the animals belong. This leads to 
confusion, because only by a discrimination of the morphology of the 
plexuses can we hope eventually to attain some concept of their origins 
and functions. 

From our own investigations and from the more detailed descrip- 
tions in the literature, we feel justified in concluding that the simple 
network type of blood vascular plexus occurs in many mammals of 
widely separated orders. Among our own material such a form of 
plexus is encountered in the Dasypodidae and in Tarsius. The great 
ant-eater {Myrmecophaga jubata), closely allied to Taviandua, likewise 
possesses a simple network type of vascular plexus in its extremities, 
but the plexiform branches are more numerous and elaborate than in 
the Dasypodidae, although, as in them, confined to forearm and leg 
(Miiller). The complete vascular bundle type occurs only in certain 
of the Xenarthra, namely the sloths and the didactyl ant-eater, and in 
certain of the Lemuroidea, namely the Lorisinae {Nycticebus, Loris, 
Perodidicus, Arctocebus). The less pronounced type of vascular bundle, 
confined to arteries and veins of forearm and leg, is limited to the 
four-toed ant-eater ( Tamandua) of the Xenarthra, and to the pangolin 
{Manis) of the PhoUdota. 

Certain species of Manis, however, may possibly possess more 
complete Aascular bundles. In Manis macrura, Hyrtl describes and 
pictures what appears to be a true vascular bundle invohing the main 
artery of the upper arm (brachial artery). The artery of the thigh 
(femoral artery), on the other hand, according to H\Ttl, is not plexi- 
form, retia mirabilia occurring only in certain branches in the leg. 
In the upper extremity of Manis lanticandata, contrary to the observa- 
tion of Hyrtl, ^Miiller states that plexuses appear only as one ap- 
proaches the distal portion of the upper arm, involving typically the 



8 bulletin: museum of comparative zoology 

median and the radial arteries. In our experience, with a specimen of 
Manw jamnica, we find the brachial and femoral vessels represented 
by single tubes (figs. 16 and 17), retia mirabilia appearing only in the 
branches of these vessels in the distal third of the arm and thigh and in 
the forearm and leg. 

Tarsius has been investigated by us because of the statement at- 
tributable originally to Vrolik (1826) and Burmeister (1846) that a 
vascular plexus is present in the lower extremity. Statements in the 
subsequent literature are confusing. Goppert (1905) claims that the 
brachial artery of Tarsius forms a tremendous retc mirabile. Fransen 
(1907) states that the external iliac artery of Tarsius gives rise to a 
plexus. Duckworth (1915) reports that the brachial artery bifurcates 
high up in the limb, but does not give rise to niia; he remarks merely 
that he observed two main vessels in the lower part of the thigh. 
Woollard (1927) does not describe a plexiform arrangement of the sub- 
clavian artery or its branches; of the external iliac artery, he says that 
as it emerges from under Poupart's ligament it breaks up into a num- 
ber of branches which can be regarded as superficial and deep. The 
exact nature of the plexus in Tarsius has never been clearly de- 
monstrated, a circumstance which leads Weber (1928) and others to 
assume a priori that the plexiform conditions in the limbs of Tarsius 
and of the lorises are identical. As we have already described, the 
brachial artery of Torsius is a single tube unaccompanied by a plexus, 
whereas both femoral artery and vein exhibit simple multiple division. 
This diffuse plexus in the thigh is, however, quite different from the 
vascular bundle encountered in this region in the lorises. 

The blood vascular bundles of the sloths, two-toed ant-eaters, 
lorises, four-toed ant-eater and pangolin are undoubtedly closely 
related in origin to the simple network type of plexus as found in 
Myrmcrophaga and the Dasypodidae. Miiller emphasizes the em- 
bryonic character of the simple network type of plexus. This he regards 
as the primitive form of blood vascular pattern; from it are derived 
both the bundle t>T)e and the usual single tubular artery. It has been 
established in man and other mammals that the various parts of the 
blood vascular system grow in the embryo as advancing plexuses of 
capillaries. Thus in the arm-buds the vascular beds are at first com- 
pletely plexiform. As the buds grow, the plexuses extend peripherally. 
Subsequently there occurs a remodehng of the original plexuses, with 
the disappearance of much of the network and the formation of 
definitive efferent and afferent pathways, which by the acquisition of 
muscular and adventitial coats become the arteries and veins. In man 



WISLOCKI AND STRAUS: BLOOD VASCULu\R BUNDLES 9 

and the majority of mammals this remolding of the embryonic plexus 
leads in the arm to the establishment of a single principal artery in 
place of the original brachial rcir. According to Midler's reasoning, in 
mammals which exhibit the arteries and \'eins in a plexiform condition 
throughout life, the resolution of the embryonic plexuses into single 
trunks does not take place. The vascular bundle type of plexus, as 
seen in the sloths and other mammals, is a highly specialized form of 
the original plexuses which have developed into a very different struc- 
ture from the embryonic pattern. 

Interesting also in this connection are the observations of Grodzinski 
(1930) upon the gradual metamorphosis of embryonic plexuses of 
blood vessels in the forelimb of | developing salamanders. He observed 
that the embryonic pattern is eliminated earlier in the proximal por- 
tion of the extremity than in the distal portion. Applying this idea to 
the material under consideration, the proximal portions of the blood 
vascular beds of both extremities in Manis, Tamandua, Myrmccophaga 
and Dasiipus have undergone complete differentiation into single 
trunks, whereas in the distal parts of these same vascular beds the 
plexiform pattern is not eliminated. In Myrmccophaga and Dasyptis 
the general embryonic pattern is retained in the distal segments, but 
in Manis and Tamandua the original network is modified into the 
highly specialized bundle type of rctc. In the sloths, two-toed ant- 
eater and lorises, on the other hand, the plexuses are not eUminated 
at all, either distally or proximally, but differentiate instead as com- 
plete vascular bundles. 

Various hypotheses have been advanced to explain the functions of 
the rcfia mirabilia, but none of these are entirely satisfactory. H\Ttl 
maintains that the so-called massive plexuses (or bundles), as found in 
sloths, are related to slow movements, and he suggests that they may 
prevent obstruction of the circulation when the musculature is in a 
state of contraction. Miiller explains that the ensheathed plexuses (or 
bundles) of the sloths facilitate the return flow of the blood from the 
limbs through the rhythmical pressure exercised upon the thin-walled 
veins by the arteries within the sheath. It has also been suggested 
that the rcfia mirabilia act as blood reservoirs. This idea is again ad- 
vanced in a recent paper by Rau and Rao (1930). 

Quite apart from any consideration of the specific functions of the 
retia mirabilia is the question of a possible correlation of the various 
types of such plexuses to modes of life. To recapitulate, our anatomical 
investigations emphasize the necessity (for the present, at least) of 
considering the sloths, the two-toed ant-eater and the lorises as pos- 



10 bulletin: museum of comparative zoology 

sessing a distinct type of rrtr viirahilc, the complete vascular bundle, 
in their extremities. From this group must be excluded the remainder 
of the lemurs and Tarsiu.s, as well as the Dasypodidae and Mi/rwrco- 
phaga. The Manidae and TcDiiandua bear a transitional position to 
the sloths, Cyclopes and the Lorisinae. To pursue our discussion, do 
the animals named in the first group (sloths, two-toed ant-eater, 
lorises) exhibit any behavior or make any characteristic use of their 
limbs which other animals do not share? This, we believe, can be 
answered in the affirmati\e. They are all arboreal, but that does not 
distinguish them from many other groups of mammals. They possess, 
however, two other characteristics which do distinguish them in a 
marked degree from all other mammals: extreme slowTiess of move- 
ment, and the ability of the musculature to maintain the animal in 
hanging or clinging positions for an extraordinary length of time. For 
the sloths this has long been known, but one of the writers (Wislocki, 
1928) has pointed out that, contrary to popular belief, the postures of 
the two sloths are quite different, the two-toed sloth alone being given 
to hanging in the way traditionally ascribed to sloths, whereas the 
usual posture of the three-toed sloth during rest is a clinging one with 
the axis of the body in the perpendicular plane, the horizontal hanging 
posture being assumed only during locomotion. Another difference 
between them is that the three-toed sloth {Bradypns) is more sluggish 
in its movements and progression than the two-toed sloth {Choice pus). 
Like the sloths, the two-toed ant-eater, Cyclopes, clings habitually and 
is extremely sluggish and deliberate in its mo\-ements. Its postin-es 
resemble strikingly those of the three-toed sloth. Its marked similarity 
in these respects to the sloth is not generally appreciated. 

Being familiar with the sloths and ant-eaters from actual observa- 
tion of li\'ing specimens, it was interesting to us to obser\'e a living 
specimen of one of the lorises, Frrodicficvs patio, which was kept in 
capti\ity for some length of time. During rest this animal's posture 
was habitually a clinging one, in the manner of the three-toed sloth. 
In locomotion along a horizontal branch it held its l)ody in the normal 
quadrupedal position, instead of in the hanging position assumed by 
the sloths and Cyclopes. ■Moreover its movements were slow and de- 
liberate, although by no means as sluggish as those of the sloths. When 
placed on the ground, Perodicticus walked slowly and awkwardly with 
legs bowed and the abdomen raised clear of the ground, contrary to 
the sloths, which are unable upon the ground to support the body upon 
the limbs. The sloths, two-toed ant-eaters and lorises constitute the 
group of animals in which complete vascular biuidles occiu" in the 
extremities. 

It is of interest to draw Tamandua and Matiis into the comparison. 



WISLOCKI AND STRAUS: BLOOD VASCULAR BUNDLES 11 

Tamandua, the four-toed ant-eater, is much more actne than either 
the sloths or the two-toed ant-eater. Nevertheless, it, too, exhibits a 
certain awkward deliherateness of movement which is reminiscent of 
the sloths. Similarly Manis, although like Tamaudun capable of 
executing fairly rapid movements in walking or climbing, is under 
most circumstances rather deliberate and slow and gi\en to the main- 
tenance of clinging postures. Tamandua and Manis are the two forms 
in which vascular Inmdles occur solely in the forearms and legs. 

Myrmecophaga, which is terrestrial and active, shows contrariwise 
scarcely a trace of the deliberate slowness of the other Myrmecopha- 
gidae {Cyclopes, Tamandua). Similarly, Dasypus, which is a terrestrial 
animal, is extremely active. Both of these animals possess only lim- 
ited, simple vascular networks. 

Finally, Tarsius and the lemurs (exclusive of the lorises) are active 
arboreal animals exhibiting none of the sluggish movements or pos- 
tural characteristics of the lorises, sloths or two-toed ant-eater. Our re- 
investigation of the blood vessels of the extremities of these animals 
places their so-called vascular plexuses in an entirely different category 
from the vascular bundles under discussion in the sloths, lorises, Cy- 
clopes, Tamandua and Manis. 

The foregoing observations on the modes of life of animals possessing 
vascular bundles in the extremities suggest, we believe, that there is a 
correlation between the blood vascular bundles and the habitual pos- 
tures and movements of these animals. The citation of these observa- 
tions by no means establishes a relationship between the vascular 
bundles and the habitual postures of these animals, but the combina- 
tion of the conditions in several widely separated mammalian groups 
suggests strongly such a possibility. The two appear to go hand in 
hand, and the similarity of the sloths, the two-toed ant-eater and the 
lorisine lemurs in these particulars can be interpreted best as instances 
of associated anatomical and functional convergences related to similar 
modes of life. Thus we return essentially to the original idea of Carlisle 
regarding a correlation betW'Cen vascular bundles and sluggishness of 
movement. 

The data which we huve been discussing are briefly summarized in 
the accompanying table. The marked similarity of sloths, Cycloprs 
and lorises in respect to the pattern of the blood vessels to the ex- 
tremities, as well as to mode of life, is apparent. The related and 
intermediate conditions in Tamandua and Manis are indicated. The 
Dasypodidae, Myrmecophaga, the Lemurinae and Tarsius differ, on 
the other hand, in both the character of their vascular patterns and 
their modes of life. There cannot, however, be said to be any correla- 
tion between simple plexiform patterns and acti\'e modes of life as 



12 



bulletin: museum of comparative zoology 



>. 















CO 



CI, 



> 



bC 



^ ^ 

^ T^ 



:3 


c 


. 




'5d 


a3 


>i 


_bE 


_> 




-3 





S- 


a 


oS 


s 




m 




00 


■bs 


X 

iq 




■5 



WD « bb 

> s > 



^ 


0) 


-^ 




-u 


c 





s 

a 


3 





£ 





u 




V 








~ 




"3 



o3 






O 

e 



03 



^ 3 









> « 

o .= 

- p 

O) -^ 

03 

03 O 

-^ ^ 

. • 53 wj 

^ iS p S 

_; ^ ^ 5 bC 

S tc > C 

OJ • _ o — 

o c c ^ O 

-c ;;; ^ .  



c 
o 

X S 
h3 O 



il 

"3 03 

- 9 
in 

-t: bc 

S bC 

S .= . 

> — «3 

O O "3 

^ .2 

a 



o 



o 
o 



> 

— S 









o 









0) « 
bC J2 






c 

3 



«— — O 



CO 

bi)-^ 
_2 o 

P c; = 



bC o 

aj 03 



o 

03 
03 

a; 
> 



c e G p 5 ■;: 
F t_ r t^ -^ iS 



? o 



c3 



X o x: o 



~ o -^ o 

0) tc 0) tc 



« « (P 



„ 1) 
" 5Q 



03 



f— wj ^-« w^ uri 

C !B C tc bIJ KT 



> p 



es ^ P -^ 



03 *--- 



P 
P c3 



> s 

1 ^ 

> QJ 

• P o 

_a) "C 



P O 

-3 o 

Hi 03 

-«^ 03 ^ .^ 

P a^ 



"P a; 
tc 



01 



> ^ O) 

^ c S 
.a S3 a> 



o 
S> tc 

tc 

CO 
0) 

> 



bC 

O) 



P 
c3 



— C -^ — G -P 

-^ p bc -a P hC 
•^ "^ Z. "^ "^ ^ 

.5f bC ^ ^ bC "5 

K .S S S -p 3 



"3 tij 



IB 
O) 



3 si 

■- o 



3 o 

s3 0) 

> 03 
■^ 03 

0; > 

J.s 

«= s 

;-< m 

sS -= 
£P -ki 

^ >> 

— -i 



0> 


CQ 


-^ 


-3 




*-r 




CJ 


'C 




Oj 


2 


-^ 


« 


bC 






p 



a; ;j 



o tc 

^ O) 



o 



ic 

o 



bij 



S3 



o 
> 

p ^ 

— o 

CO «« 

^ 73 

■^ P 

s °* 

^ s 

i: =^ 

^ u 

3 1^ 

o C^ 

CO CL 

c3 3 

-T3 O 

U 03 



> 



c3 



O) 



X 

a 



33 



As i:  



o S 

> 0) 

p ^ 

-g Oj 

I- 

O) -<-' 

P CO 

P ^ 
p tiO 
1- CO 

O « 
"3 •* 

o 03 



0) 

_> 

03 






S3 

01 



3 






'> 
-5 

o> 



"S "^ bl «i -P 

•5 Sg ^ & g 

G a^ G b 

S > ^ .P 2 

^ e P 03 _Q 

^ e -P -^ S 

"3 £ bC i 'S 

>-, o IS P "£ 

— ^ _G :i3 « 5 r-, 

-fr -^ — ^ " • C 

■- Sf ;:; X bc aj .3 



3 

s 

aT 








e 






!3i 






e 












a. 



1 




g 


-B 




S 

v. 





a. 






•*-:> C: 


ti 


-c e 




S i i5 






O 
•^ 



1) 



3 

as 



WISLOCKI AND STRAUS: BLOOD VASCULAR BUNDLES 13 

postulated by Hyrtl. A brief consideration of various actixe groups of 
mammals would amply demonstrate this. As has been said, it is likely 
that the simple networks are a widespread, less differentiated, primi- 
tive condition. Consequently, in the order of the Xenarthra, the vessels 
to the extremities are primitive in the Dasypodidae and Myrmeco- 
phaga, partially specialized in Tamandua, and highly specialized in 
Cyclopes, Bradypus and Choh'pus. 

It is of interest to note that the striated musculature of the Brady- 
podidae is uniformly of an exceedingly dark red variety (Wislocki, 
1928). We have also noted the same circumstance in unrelated Pero- 
didicus. In the Dasypodidae, on the other hand, the striated muscu- 
lature is mixed pale red and white. In the other animals under con- 
sideration the character of the musculature in freshly killed specimens 
is unknown. These findings, nevertheless, suggest that dark red col- 
oration possibly characterizes the musculature of those mammals in 
which there is habitual sluggishness of movement. 

Suvimary 

At least two types of rcfia mirahiJia or blood vascular plexuses are 
found in the extremities of mammals. The one type comprising the 
majority of retia, which are probably primitive, consists of simple 
plexiform anastomoses. The other type, consisting of highly organized 
vascular bundles, is probably a very specialized structure. 

Typical vascular bundles involving the whole course of the brachial 
and femoral arteries and their ultimate major divisions are encoun- 
tered in the two-toed and three-toed sloths {Choh'pus and Bradypus), 
the two-toed ant-eater {Cyclopes) and in the Lorisinae. Typical vas- 
cular bundles involving only the \essels of the forearm and leg are 
encountered in the pangolin {Manis) and the four-toed ant-eater {Ta- 
mandua). The great ant-eater {Myrmecophaga) and the armadillos, 
on the other hand, possess refia mirabilia of the simple anastomosing 
type. 

Tarsius possesses retia only in its lower extremities, involving each 
femoral artery. The latter, however, does not form a typical vascular 
bundle, as in the sloths and lorises, but breaks up completely into a 
set of small radiating arteries. 

The typical vascular bundle occurs only in sluggish arboreal animals. 
The sloths and lorises, in which the bundles are typically seen, are 
phylogenetically only distantly related. This circumstance suggests 
that these vascular bundles have developed in these forms as adaptive 
convergences to their similar modes of life. 



14 bulletin: museum of comparative zoology 



BIBLIOGRAPHY 

Baer, K. E. von 

1823. Beitrag zur Kenntniss vom Bau des dreizehigen Faulthiers. 

Deutsch. Arch. f. Physiol., 8, p. 354. 
1835a. Ueber das Gefasssystem des Braunfisches. Xova Acta Acad. 

Caes. Leop. Carol, naturae curiosorum, 17. 
1835b. Ueber die Geflechte in welche sich einige grossere Schlagadern der 

Siiugethiere friih aiiflosen. Mem. presente a I'Acad. imperials 

des" sciences de Saint-Petersbourg, 2. 

BURMEISTER, H. 

1846. Beitrage zur niiheren Kenntniss der Gattung Tarsius. 140 pp., 
Berlin. 
Carlisle, A. 

1800. An account of a peculiarity in the distribution of the arteries sent 

to the limbs of slow moving animals. Phil. Trans. London, p. 601. 

1804. Continuation of an account of a peculiar arrangement in the 

arteries distributed on the muscles of slow moving animals. Phil. 

Trans. London, p. 17. 

Chapman, H. C. 

1874. Rete Mirabile in Bradypus Didactylus. Proc. Acad. Nat. Sci. 
Phila., p. 95. 
Duckworth, W. L. H. 

1915. Morphology and anthropologj'. Cambridge, 2d edition, 1, p. 111. 
Fransen, J. W. P. 

1907. Le systeme vasculaire abdominal et pelvien des Primates. Petrus 
Camper, Nederl. Bijdr. Anatomie, 4, pp. 215, 487. 
Gegenbatjr, C. 

1901. Vergleichende Anatomie der Wirbeltiere, 2. 
Goppert, E. 

1905. Mammalia' Arteria axillaris imd brachialis. Bronn's Klassen u. 
Ordnungen des Thier-Reichs, 6, pt. 5, p. 1,221. 
Grodzinski, Z. 

1930. Die Blutgefassenentwicklung in der vorderen Extremitat bei 
Amblystoma mexicanum Cope. Bull. I'Acad. Polonaise Sc. et 
Lettres, CI. Sc. Math, et Nat., Ser. B, Sc. Nat., 2, p. 247. 
Hyrtl, J. 

1853. Das arterielle Gefasssystem der Edentaten. Denkschr. d. Kais. 

Akad. Wissensch. Wien, 6, p. 21. 
1864. Neue Wundernetze und Geflechte bei Vogeln und Siiugethieren. 
Denkschr. Kais. Akad. Wissensch. Wien, 22, p. 113. 
Huxley, T. H. 

1864. On the angwantibo (Arctocebus calabarensis Gray) of old Calabar. 
Proc. Zool. Soc. London, p. 314. 



WISLOCKI AND STRAUS: BLOOD VASCUL.\R BUNDLES 15 

Miller, E. 

1905. Beitiage zur Morphologie ties Gefiisssystems. II. Die Armar- 
terien der Saugetiere. Anat. Hefte, 1, 27, p. 122. 
Rapp, W. V. 

1852. Anatomische X'ntersuchungen fiber die Edentaten. Tubingen. 
Rau and Rao 

1930. (Arteries of Loris lyddekerianus.) Jour. Mysore Univ., 4. (Seen 
only in abstract in Sc. Prog., No. 99, January, 1931, p. 412.) 
Vrolik, W. 

1826. Disquisitio anatomica-physiologica de peculiari arteriarum ex- 
tremitatum in nonnullis animalibus dispositione. Amsteledami, 
C. G. Sulpke. 
Weber, M. 

1928. Die Saugetiere. 2d edition, 2 vols., Jena. 
WiSLOCKI, G. B. 

1928. Observations on the gross and microscopic anatomy of the sloths 
{Bradypus griseus griseus Gray and Choloepus hoffmanni Peters). 
Journ. Morphol. and Physiol., 46, p. 317. 
WOOLLARD, H. H. 

1925. The anatomy of Tarsius spectrum. Proc. Zool. Soc. London, 
pt. 3, p. 1,071. 

ZtJCKERKANDL, E. 

1896a. Zur Anatomie und Entwicklungsgeschichte der Arterien des 

Vorderarmes. Anat. Hefte, 1, 5, p. 196. 
1895b. Zur Anatomie und Entwicklungsgeschichte der Arterien des 

Unterschenkels und des Fusses. Anat. Hefte, 1, 5, p. 216. 



EXPLANATION OF PLATES 



PLATE 1 



WiSLocKi AND Straus. — Blood Vascular Bundles 



PLATE 1 

Fig. 1. Typical cross-section of the upper arm bundle of the three-toed sloth 
{Bradypus griseus) showing the brachial artery surrounded by a rete 
tnirabile of smaller arteries and veins. The entire vascular bundle is en- 
closed and held together by a dense connective tissue sheath. The vessels 
are filled with an India ink injection mass. X 20. 

Fig. 2. Typical cross-section of the thigh bundle of the three-toed sloth 
{Brndypus griseus) showing the femoral artery and its associated rete 
mirahile of smaller arteries and veins. The entire structure is held to- 
gether by dense fibrous tissue. Vascular injection with India ink. X 20. 

Fig. 3. Typical cross-section of the upper arm bimdle of the didactyl ant- 
eater {Cyclopes didactylus) showing an association of a rete mirahile with 
two stout arteries. X 20. 

Fig. 4. Typical cross-section of the thigli bundle of the didactyl ant-eater 
(Cyclopes didactylus) showing a similar association of a rete mirahile 
of small ai-teries and veins with two larger arteries. X 20. 
The four figures on this plate illustrate the close similarity of the condi- 
tions found in the upper arm and thigh blood vessels of sloths and di- 
dactyl ant-eater. Each of the sections is magnified twenty times, the 
upper arm bundles of the sloths being absolutely larger than those of the 
smaller didactyl ant-eater. The sloth material was well preserved and 
shows good detail; the material from the didactyl ant-eater was taken 
from an old alcohol-fixed museum specimen. Nevertheless the latter is 
well enough preserved to show the similarity of the vascular bundles in 
the two forms. 



BULL. MUS. COMP. ZOOL. 



WiSLOCKi AND Straus. Plate 1 




^•r^ 



>A 



i^ 


















s 
1 






V^>; 



■^ 



V 













HELlOTYPt CO. BOSTOS 



PLATE 2 



WiSLOCKi AND Straus. — Blood Vascular Bundles 



PLATE 2 

Fig. 5. Typical cross-section of the upper arm bundle of a slow lemur, Pero- 
didicus potto, showing the brachial artery surrounded by a rete mirabile 
composed of smaller arteries and veins, the whole bound together by 
dense connective tissue. X 50. 

Fig. 6. Typical cross-section of the thigh bundle of Perodicticus potto, show- 
ing two main arteries surrounded by a rete mirabile of smaller arteries 
and veins. X 50. 

Fig. 7. Typical cross-section of the upper arm bundle of another one of the 
lorises (Nycticebus borneanus) showing the brachial artery in association 
with a rete mirabile. X 50. 

Fig. 8. Typical cross-section of the thigh bundle of Nycticebus borneanus. 
X 50. 

These four figures, aU at the same magnification, show the tjTiical character 
of the vascular bundles of upper arm and thigh in the lorises. Notice that 
they are practically identical with the vascular bundles involving the 
brachial and femoral arteries o<" the sloths and didactyl ant-eater. The 
material from Perodicticus is well preserved from a fresh specimen, that 
from Nycticebus is from an old alcoholic museum specimen. 



BULL. MUS. COMP. 200L, 



WisLOCKi AND Straus. Plate 2 
















rV?" -i 



-« 









HELlOTYPt CO. BOSTON 



PLATE 3 



WiSLOCKi AND Straus. — Blood Vascular Bundles 



PLATE 3 

Fig. 9. IVpical cross-section of the brachial artery of one of the Lemvrinae 
{Lemur vartegatus). Note that the artery is relatively large and that it 
is not associated with a rete mirabile. It is accompanied by a brachial 
vein of about equal size. X 50. 

Fig. 10. Typical cross-section of the femoral artery and vein of Lemur 
vartegatus. Note that artery and vein are of about equal size and rela- 
tively lai'ge and that they are not associated with a rete mirabile. X 50. 

Fig. 11. Typical cross-section of the brachial artery of Tarsius fraterculus 
showing a relatively large artery which is not accompanied by a rele 
mirabile. The tissues from this Tarsius are jjoorly preserved, but show- 
nevertheless the topography under discussion. X 50. 

Fig. 12. Typical cross-section through the thigh of Tarsms /ra^e?TwZtis. Note 
that there is no main artery. Observe also that there is no vascular bundle. 
Instead there are scattered clusters of arteries and veins of about equal 
size and number which are not ensheathed by dense connective tissue 
capsules. The femoral plexus in Tarsiua, produced by radial branching, 
is morphologically dissimilar from the vascular bundles of the sloths, di- 
dactyl ant-eater and the lorises. X 50. 



BULL. MUS. COMP. 200L. 



WisLOCKi AND Straus. Plate 3 














»-i ^- ♦-~- T-.^ > , . .., 



0^ 




HELIOTYPt CO. BOSTON 



PLATE 4 



WiSLOCKi AND Straus. — Blood Vascular Bundles 



PLATE 4 

Fig. 13. Typical cross-section of the brachial artery of the four-toed ant- 
eater {Tamandua tetradaclyla) . Note the relatively large size of the 
artery and that it is not associated with a rete mirahile. X 20. 

Fig. 14. Typical cross-section of the femoral artery of Tamandua tetradaclyla. 
This vessel also is not associated with the formation of a vascular bundle, 
differing in this respect from the didactyl ant-eater and the sloths. X 50. 

Fig. 15. Cross-section of the radial bundle of Taviandua tetradaclyla, showing 
the typical vascular bundle formation in the arteries of the forearm. X 20. 

Fig. 16. Typical cross-section of the brachial artery of Manis javanica, show- 
ing the single, large artery devoid of a vascular bundle. X 50. 

Fig. 17. Typical cross-section of the femoral arterj^ of Man Js javanica, show- 
ing a single larger artery in the absence of a vascular bundle. X 50. 

Fig. 18. Cross-section of the tibial vascular bundle in Manis javanica, illus- 
trating that the arteries of the leg give rise to typical vascular bundles. 
X 20. 



BULL. MUS. COMP. 200L. 



WiSLOCKi AND Straus. Plate 4 




HELlOTYPt CO. BOSTON 



JUL 1 1932 



3)^7 

Bulletin of the Museum of Compa,ratl\e Zcoiogy 

AT HARVARD COl'l'eGE 
Vol. LXXIV, No. 2 



A STUDY OF THE OSTEOLOGY OF ALLIGATOR 
PRENASALIS (LOOMIS) 



By Charles C. Mook 



With Three Plates 



CAMBRIDGE, MASS., U. S. A. 

PRINTED FOR THE MUSEUM 

July, 1932 



PUBLICATIONS 
OF THE 

MUSEUM OF COMPARATIVE ZOOLOGY 
AT HARVARD COLLEGE 



There have been pubUshed of the Bulletin Vols. I to LXV, LXVII- 
LXXIII;of the Memoirs Vols. I to LL . 

The Bulletin and Memoirs are devoted to the publication of 
original work by the Officers of the Museum, of investigations carried 
on by students and others in the different Laboratories of Natural 
History, and of work by specialists based upon the Museum Collec- 
tions and Exploration. 

These publications are issued in numbers at irregular intervals. 
Each number of the Bulletin and of the Memoirs may be sold sepa- 
rately. A price list of the publications of the Museum will be sent on 
application to the Director of the Museum of Comparative Zoology, 
Cambridge, Massachusetts. 



Bulletin of the Museum of Comparative Zoology 

AT HARVARD COLLEGE 

Vol. LXXIV, No. 2 



A STUDY OF THE OSTEOLOGY OF ALLIGATOR 
PRENASALIS (LOOMIS) 



By Charles C. Mook 



With Three Plates 



CAMBRIDGE, MASS., U. S. A. 

PRINTED FOR THE MUSEUM 
July, 1932 



No. 2. — A Study of the Osteology of Alligator prenasalis (Loomis) 
By Charles C. ISIook^ 

Introduction 

In prospecting for fossil mammals in the Oligocene beds of South 
Dakota in 1925, Messrs. Hugo and Eric Schlaikjer discovered some 
well preserved crocodilian remains. These remains were taken up in 
the field and shipped to the Museum of Comparative Zoology at Cam- 
bridge, Massachusetts, and were presented to that institution by Dr. 
Thomas Barbour. The locality of this occurrence of fossil crocodiles 
is fifteen miles southwest of Scenic, South Dakota, and the level is 
the Titanotherium Beds of the White River Formation, of Lower 
Oligocene age. The bones were found one hundred and thirty-four 
feet above the Cretaceous-White River contact, and forty-eight feet 
below the base of the lower Oreodon Beds. 

The specimens consist of two individuals, both well preserved, but 
neither of them perfect. Between them they exhibit the characters of 
nearly all parts of the skeleton. These fossils were noted in a brief 
article by Dr. Barbour in Copeia, No. 151, pp. 109-111, 1926, and 
were correctly identified by him as belonging to Alligator preyiasalis 
(Loomis). The writer is indebted to Dr. Barbour for the opportunity 
of stud^-ing and describing this material, and to Prof. Henry Fairfield 
Osborn, of the American Museum of Natural History, who assigned a 
grant from the Osborn Research Fund, which defrayed the cost of the 
dra\Anngs. The photographs were made at the Museum of Compara- 
tive Zoology, and the drawings were made by Mr. John C. Germann 
and Mrs. Louise W. Germann, of the American Museum of Natural 
History. 

General Form of the Skull 

Alligatoroid, the lateral margins are regular, converging very 
gradually forward in practically straight lines from the posterior tip 
of the quadratojugal to the level of the third maxillary teeth; from 
this point forward they curve inward sharply to meet at the mid-line 
at the tip of the snout. The convergence of the posterior portions is 
considerably greater than in the modern alligator. 

This convergence of the borders is correlated with a relatively 
greater breadth of the posterior portion of the skull in the specimens 

1 Contributions to the Osteology, Affinities, and Distribution of the Crocodilia No. 22. 



20 



bulletin: museum of comparative zoology 



described than in the Florida alUgator. The difference in breadth 
between the anterior and posterior portions of the skull is less in A. 
mississippiensis than in the South Dakota specimens. This is equally 




Fig. 1. Alligator prenasalis (Loomis). 
natural size. Mus. Comp. Zool. No. 1,015. 



Skull, superior view. One-half 



true whether juvenile or adult specimens of A. mississippiensis are 
used for comparison. 

In cross profile the skull is higher than in cither of the living species 
of Alligator or in A.thoinsoni. This is especially true along the mid-line 



mook: osteology of alligator prenasalis 



21 



of the snout. The nasal bones are elevated above the premaxillaries. 
The interorbital region of the frontal slopes gradually to the surface 
of the nasals, without an abrupt descent as in other alligators and in 




Fig. 2. Alligator preriasalis (Loomis). 
size. Mus. Comp. Zool. Xo. 1,015. 



Skull, palatal view. One-half natural 



the living Jacare. This fact is the more notable because it is apparent 
in spite of slight crushing in a direction that would tend to obscure it. 
I' The cranial table is larger in every way than in the Florida or 
Chinese alligators. It is slightly longer, and is considerably broader, 



22 bulletin: museum of comparative zoology 

especially along the posterior border. Like the lateral borders of the 
skull as a whole, the lateral borders of the cranial table converge more 
sharply forward than in the living alligators. The space between the 
orbits is narrower and that between the supratemporal fenestrae 
wider than in the living alligator. 

Fenestrae of the Skull 

The supratemporal fenestrae are larger than in the living alligators, 
occupying more of the area of the cranial table. They are semicircular 
posteriorly; their external borders are nearly straight antero-posterior 
lines ; their antero-internal borders bend sharply forward and outward 
from opposite the centers of the fenestrae, giving the latter an asym- 
metric appearance, with pointed ends directed forward and outward. 

The orbits are similar in shape to those of the living alligators, but 
are relatively larger. 

The external narial opening is characteristic. It is very broad. In 
the living alligators the nasal bones extend forward and join the pre- 
maxillaries in front of the narial aperture, thus separating the latter 
into two distinct orifices, at least at the surface. In A. thomsoni the 
nasals extend forward into the aperture, but do not reach the pre- 
maxillaries in front of it. This may be due to incomplete preservation, 
as the tip of the nasal projection in the living forms is very delicate. 
In the specimens now described the nasals project forward but do not 
reach the anterior cross-bar of the maxillaries. The anterior border of 
the aperture is low and the lateral and anterior premaxillary surfaces 
slope gradually up to the exterior making the aperture shallow except 
in its center. The low anterior border gives the aperture the ap- 
pearance of being directed forward. This form of the aperture cor- 
responds exactly with that in the type of Alligator prcnasalis Loomis, 
and difl'ers from all other known species of crocodilians. 

The lateral temporal fenestra is similar in shape to that in modern 
alligators, but is relatively larger. 

On the palate the premaxillary fenestra resembles in size and shape 
that of the living alligators. 

The palatine fenestra is distorted on both sides in both specimens. 
On analyzing the effects of this distortion it becomes apparent that the 
palatine fenestrae were shorter and broader, and were nearer together 
than in the Recent alligators. Their external borders were also farther 
from the tooth row. In other words the fenestrae are situated nearer 
the mid-line than in the Recent alligators. 



mook: osteology of alligator prenasalis 



23 



Mandible 

The lower jaws are more massively constructed than are those of 
the Recent alligators. The symphysis is longer, evidently being op- 
posite eight teeth on each side. (The dental borders of No. 1015 are 




Fig. 3. Alligator prenasalis (Loomis). Lower jaws, superior view. One-half 
natural size. Mus. Comp. Zool. No. 1,015. 



24 bulletin: museum of comparative zoology 

incomplete, and those of 1014 are covered, so this point cannot be 
definitely determined at present.) The splenials definitely enter the 
sjonphysis and form its posterior border. 

The shaft of each ramus is broad and moderately high. 

The external mandibular fenestra is not well preserved in either 
specimen. In the right ramus of 1015, however, its outlines are dis- 
tinct enough to indicate that its size was small. 

Dentition 

In No. 1014 the dentition is partially obscured because the lower 
jaws are firmly attached to the skull. In No. 1015 the alveolar borders 
are somewhat damaged. No positive statement can be made, there- 
fore, regarding the dental formula. 

In 1014 four premaxillary teeth are present on each side, but these 
do not exactly correspond in position wdth each other, or wnth the 
teeth of 1015. They indicate that the complete jaw had five premaxil- 
lary teeth on each side. In 1014 the left side has three premaxillary 
teeth and spaces for two alveoli between the anterior of the three and 
the mid-line. The right premaxillary has four teeth, and the alveolus 
of a fifth. Two of the teeth in the right premaxillary correspond with 
an incomplete border in the left.. It appears, therefore, that five is 
the characteristic number of teeth for each premaxillary. 

In No. 1015 the right maxillary contains twelve actual teeth and 
an incomplete posterior groove that must have held at least two, and 
possibly three more teeth. There is a space between the anterior 
maxillary tooth and the posterior premaxillary tooth that may have 
lodged a small tooth. The left maxillary contains five actual teeth, 
and the border is incomplete, so the original number cannot be de- 
termined. The number of maxillary teeth on each side indicated by 
this specimen is sixteen. 

In No. 1014 ten teeth are present in the right maxilliary, with clear 
indication of two more. The posterior portion of the alveolar border 
is not clearly ^dsible, but three more teeth in this region seems hkely. 

The left maxillary contains fourteen teeth in a continuous series 
from the premaxillary suture backward. The posterior end of the 
alveolar border is not visible, but indications are present that another 
tooth was lodged back of No. 14. On the whole the evidence points 
toward fifteen teeth in each maxillary. This added to the five pre- 
maxillarv teeth noted above, makes twentv teeth on each side of the 
upper jaw. 

The lower teeth of No. 1014 are entirely invisible in the specimen. 



mook: osteology of alligator prenasalis 25 

In No. 1015 ten teeth are preserved on the right side and fourteen 
on the left. The anterior border of the lower jaw is incomplete, making 
it difficult to estimate the number of teeth opposite the s^Tnphysis. 
The posterior dental borders are also more or less incomplete. Judging 
from the size, shape, and spacing of the visible teeth and alveoli the 
number of mandibular teeth on each side was twenty or twenty-one, 
and more probably twenty-one than twenty. Probably eight teeth 
were opposite the symphysis on each side. 

The teeth themselves are very stoutly constructed. In the premaxil- 
laries the first, second and fifth teeth are small, the third is of moderate 
size and the fourth is the largest. The premaxillary teeth are about 
evenly spaced with each other. 

In the maxillarles the first two teeth are small and are very close 
together. The third is very much larger, while the fourth is almost 
gigantic in size. In the interlocked position of the jaws in No. 1014 
the upper third maxillary tooth projects downward as far as the lower 
border of the mandible. The fourth maxillary tooth projects consid- 
erably below this level. The volume of the third and fourth maxillary 
teeth may be estimated at about eight times that of correspondmg 
teeth in a Florida alligator of approximately the same size. All of 
the teeth thus far noted are asymmetrically conical, with relatively 
long crowns. The next tooth (fifth maxillary) is small. It has a short 
subconical crown. The sixth and seventh, and to a certain extent the 
eighth maxillary teeth are very small. Their crowTis are very short 
vertically. They are much longer antero-posteriorly than they are 
broad laterally, at the same time they are sharp pointed. 

Posterior to the eighth the maxillary teeth are of moderate size, 
with long roots and short blunt crowns, very similar to those of Allog- 
nathosvchvs of the Paleocene and Eocene. 

The surfaces of the crowTis of all the teeth have very fine vertical 
striations. In the anterior teeth these are so faint that they do not 
interfere with a smooth or polished appearance of the teeth. Posterior 
to the fourth maxillary tooth, however, they are distinct enough to 
give a satiny appearance to the crowns. 

The order in size of some of these teeth will be: 

largest 4th max-illary 

second largest 3rd maxillary 

third largest 4th premaxillary 

The mandibular teeth are not visible in No. 1014 and are incom- 
plete in No. 1015. Apparently there was a medium to small-sized 
tooth near the symphysis, then two slightly larger, the fourth con- 



26 bulletin: museum of comparative zoology 

siderably larger, then eight or nine small teeth in the anterior sag of 
the dentary, then a much larger tooth, posterior to which the teeth 
gradually decreased in size. The total number of lower teeth is un- 
certain. It was probably twenty-one in each ramus. 

The Bones of the Skull 

PreinaxiUaries. The premaxillaries are short and broad. Their 
sutures with the nasals extend back as far as the level of the large 
fourth maxillary tooth. On the borders of the skull the premaxillo- 
maxillary sutures lie opposite the posterior third of the external narial 
aperture. On the palate the suture is covered in No. 1014 and it is 
obscure in No. 1015. It is certain however, that it did not extend back 
of the level of the second maxillary tooth. The posterior premaxillary 
processes on the dorsal surface are not sharply chfferentiated from the 
principal mass of the bone. The premaxillary teeth are subequally 
spaced from each other. The rim that borders the narial aperture is 
elevated at the postero-external edges, leaving a very narrow space 
betM'een their bases and the premaxillo-maxillary suture. The an- 
terior and antero-external portions of the narial rim are depressed. 

Maxillarics. The maxillaries are short and broad. They occupy 
somewhat less of the total breadth of the skull on the snout than in 
the Florida alligator, the nasals being somewhat broader. Their 
sutures with the nasals are nearly straight lines that are almost parallel, 
converging only very slightly forward. These sutures are relatively 
much shorter than in the Florida aUigator. The sutures with the pre- 
frontals are about the same length in A. mississippiensis. The sutures 
with the lachrymals are similar in form to those of the latter species. 

The maxillaries occupy most of the palate. The maxillo-palatine 
suture is somewhat obscure in both specimens, but it appears to curve 
forward and inward regularly from the anterior or antero-internal 
border of the palatine fenestra on one side to the mid-hne near the 
level of the seventh maxillary tooth and similarly outward and back- 
ward to the other side. The suture with the ectopterygoid is covered 
by the mandible in No. 1014 and is somewhat obscured in No. 1015. 
It is evident, however, that it is somewhat longer than in A. viissis- 
nippiensis. 

The first ten maxillary teeth are variously spaced on the opposite 
sides of the two specimens, and have separate alveoli. The posterior 
maxillary teeth are close together and are lodged in a common alveolar 
groove. 

Nasals. These bones are broad, occupying a slightly greater portion 



mook: osteology of alligator prenasalis 27 

of the total breadth of the snout than in A. mississipjnensls or A. 
sincnsc. The process entering the narial aperture is broader and blunter 
than in the Florida alligator. The breadth of the portion of the pos- 
terior border of the narial aperture occupied by the nasals is 18 mm. 
in No. 1014, compared with 13^ mm. in a Florida alligator skull of 
the same length (Am. Mus. No. 7122). 

The nasals extend back as far as the eleventh maxillary tooth. 
In A. sinensis (Am. Mus. No. 23,900) they extend to the level of the 
eleventh maxillary tooth, and m A.mississippiensis they extend back 
to the twelfth to fifteenth maxillary teeth (Am. Mus. Nos. 7,122 and 
9,043 respectively) . It is possible that this character may be correlated 
somewhat with the indi^•idual age of the specimen. 

The short contact between the nasals and the maxillaries has been 
mentioned above. The contact with the prefrontals is long, measuring 
25 mm. as against 20 in a Florida alligator of the same skull length. 

Lachrymah. In No. 1015 the lachrymal borders are not sufficiently 
preserved to warrant description. The lachrymals are well preserved 
in No. 1014 and the following description is based upon that specimen. 
The bone is somewhat larger than in ^1. viississippiensis, and extends 
forward to the level of the eighth premaxillary tooth instead of to 
the tenth or eleventh as in ^4. viississippiensis or the ninth as in A. 
sinense (Am. Mus. 23,900 juv.) The suture with the prefrontal is 
longer than in the living American species. The wedge of the bone 
h-ing between the inferior border of the orbit and the jugal bone is 
narrower than in the latter species. As in both species of living alli- 
gators the lachrymals have no contacts with the nasals, and their 
longitudinal axes, or axes of greatest length, are parallel with the 
longitudinal axis of the skull. 

Prefrontals. These bones are extremely long and slender. In this 
respect they greatly exceed ^. viississippiensis in which the prefrontals 
are long and slender. Their sutures with the maxillaries are less than 
one-half as long as their sutures with the nasals, and about one-fourth 
the length of their contact wHth the frontal. 

Frontal. The anterior process of the frontal is long and slender. 
It is relatively longer than in an A.mississippiensis specimen of the 
same skull length (Am. Mus. 7,122) but is relatively shorter than in an 
older A. viississippiensis skull (Am. Mus. No. 9,043.) Its relative 
length is therefore to be correlated with the variation in snout length 
depending upon the individual age of the specimen. The interorbital 
plate is narrow, and its edges are not uprolled in either specimen. 
The expansion posterior to the orbits is very slight. 



28 bulletin: museum of comparative zoology 

Postorbitals. These bones are not especially characteristic. The 
differences between them and the postorbitals of a given A. 7nissis- 
svp2)ie7i.sis skull are less than the differences between various skulls of 
A. mississippiensis. 

Squamosals. The squamosals, like the postorbitals, are verj' similar 
to those of the Recent Florida alligator 

Parietal. There is nothing particularly distinctive about the parie- 
tal. As in both Recent species it occupies the central region of the 
posterior border of the cranial table, confining the supraoccipital 
to the posterior surface of the skull. 

Supraoccijntal. As in the living alligator the supraoccipital is con- 
fined entirely to the posterior surface of the skull, forming no part of 
the posterior border of the cranial table. Its boundaries cannot be 
made out in No. 1015 and are somewhat obscure in No. 1014. Its 
breadth is three-sevenths of the breadth of the postoccipital surface. 
Its height is uncertain, but it is probably about one-half of the dis- 
tance from the cranial table to the superior border of the foramen 
magnum; this is somewhat less than in A. viississippien.sis. 

Exoccipitals, Basioccipital, and Basisphcnoid. These bones differ 
in no essential characters from those of modern alligators, and need no 
special description. 

Quadrates. The quadrates are very broad and stout compared 
with those of the living alligators. The breadth of the glenoid surface is 
about four-thirds of that of a Florida alligator of the same skull length. 

Qtiadrato-jngals. These bones are somewhat stouter than in the 
Hving alligators. The flanges of bone separating the quadrates from 
the postero-internal borders of the lateral temporal fenestrae are 
unusually broad. The quadrato-jugals also occupy sHghtly more of 
the external border of this fenestra than in the living alligators. 

Jugals. The jugals are somewhat longer and more slender than in 
the living alligators. Their anterior boundaries are at the level of the 
tenth maxillary tooth rather than the eleventh or twelfth as in A. 
mississippiensis In A. sinense they are over the ninth or tenth maxil- 
lary tooth. Individual age undoubtedly influences this character 
somewhat. 

Palatines. These bones extend forward to the level of the eighth 
maxillary tooth. The sutures with the maxillaries curve forward from 
the palatine fenestrae to meet at the mid-line, contrasting with A. 
mi.ssissip2)ie7isis apd A. thomsoni, but resembling more A. sinense. 
Their sutures with the pterygoids are slightly farther forward than in 
A. viississipjnensis. 



mook: osteology of alligator prenasalis 29 

Pterygoids. The pterygoids are somewhat crushed in both speci- 
mens. It is clear, however, that they are short antero-posteriorly, 
and that they occupy small portions of the internal borders of the 
palatine fenestrae as well as the posterior borders. Crushing has 
slightly modified the borders and position of the internal narial aper- 
ture, but there can be no doubt that it was situated slightly farther 
forward than in A. thowsoni, A. sinensc and A. mississippiensis. In 
this respect it is intermeditae V)etween these forms and AUognatho- 
suchus mooki Simpson, from the Paleocene. The total breadth of the 
19 pterygoids is much greater than in modern forms. 

Ectoptrrygoids. These bones differ in no essential characters from 
those of li\dng alligators except that they are more stout. 

Mandible 

The mandible is very broad, both at the symphysis and at the 
posterior end. The symphysis is considerably longer than in the living 
alligators, extending backward to the level of the eighth or ninth 
maxillary tooth. The exact level cannot be determined as the teeth 
are not visible in No. 1014, and in No. 1015 are incomplete. The in- 
dividual rami are thick laterally, and they diverge rapidly in the 
posterior direction. Consequent upon these facts the symphysial 
surface, viewed from above, is much larger than in living alligators. 
The external mandibular fenestra is rather short, being about one- 
eighth as long as the entire jaw, compared with fifteen to eighteen 
per cent in both species of living alligators. The height of the jaw is 
slightly less than in modern alligators. 

The individual bones of the mandible are not especially character- 
istic. It may be noted that the splenials enter and form a small part 
of the symphysis as in the living alligators. The articular bone is 
broader than in the Recent forms. 

Scapula and Coracoid 

The scapula of the left side is fairly well preserved; that of the right 
side being fragmentary. The length of the left scapula compared to 
the breadth of the shaft and of the distal end is small compared with 
C. americanus and A. viississippiensis. In other words the scapula is 
unusually short. The surface for articulation with the coracoid is 
more oblique than in the living Florida alligator and crocodiles. This 
makes a somewhat different angulation between the scapula and the 
coracoid than in the modern forms. The process on the antero-external 



so 



bulletin: museum of comparative zoology 



border is unusually prominent but this may have been accentuated 
b}^ crushing. 

Both coracoids are well preserved. Like the scapulae they are short 
and stout. Otherwise they resemble the coracoids of the modern alli- 







Fig. 4. Alligator prenasalis (Loomis). A, Left scapula, external view; B, 
left coracoid, external view; C, left ilium, external view; D, left pubis, external 
view. One-half natural size. Mus. Comp. Zool. No. 1,014. 




Fig. 5. Alligator prenasalis (Loomis). A, right humerus, anterior view; 
B, right ulna, external view; C, right radius, external view; D, left manus, 
superior view. One-half natural size. All views Mus. Comp. Zool. No. 1,014, 
except one ungual phalanx, which is No. 1,015. 

gator in form. The coracoid foramen is somewhat nearer the mid- 
line than in the Florida alligator and much nearer than in the Florida 
crocodile. As noted in the discussion of the scapula the angulation 
between the scapula and the coracoid is somewhat different from that 



mook: osteology of alligator prenasalis 



31 



of modern crocodilians. The long axes of the two bones are somewhat 
more nearly parallel than in A. mississippiensis, and much more nearly 
parallel than in C. amcricanus. 

Humerus 

Both humeri are well preserved and the characters noted may be 
observed equally well on both of them. The humerus is longer in 
proportion to the breadth of the proximal end than in the Florida 
alligator and the Puerco Allognathosuchus but is shorter than in 
Croeodilus americanus. The same is true with respect to the distal 
end and the circumference of the shaft. The tip of the deltoid crest is 
relatively nearer the center of the bone than in C. americanus and 
A. mississippiensis, but less near than in Allognathosuchus mooki. 
The crest projects farther from the anterior surface of the bone than 
in any of these species. 



Measurements and Ratios 



a 

*^ C 
^ O 

£•- 

n S 

CD .5 

^^ 



IS 

Length, median 259 

Length, maximum 274 

Breadth prox. end 72 

Breadth distal end 66 

Breadth shaft minimmn 30 

Circumference of shaft minimum 103 

Index: Circumference over maximum length 376 
A Center of proximal end to center of deltoid 

crest 74 

B Center of deltoid crest to center of distal end 215 

Ratio A over B 349 

Ratio, length median over length of femur 

median 794 

Ratio, length median over length of ulna 

median 1497 

Ratio, length median over length of radius 

median 1762 



^ s 

J » 

^ 3 

00 u 
<D § 
«■£ 

3 a 

Is 

130 
137 
44 
39 
18 
63 
459 



42 

99 

424 



1326 
1547 



B.-2 

3 « 

£.1 

00 g 

3.2 

S3 



158 
162 
50 
47 
21 
77 
475 

53 
128 

414 



1254 
1436 



a: 



o e 
Ng 

aa, 
Si^ 
05 
Ob 
. 01 

§5 

100 

103 

28 

27 

15 

44 

427 



35 

75 

493 

798 

1282 

1492 



32 bulletin: museum of compakative zoology 

Radius and Ulna 

The radius and ulna are present on both sides and are well preserved. 
They resemble the corresponding bones of A. mississippiensis and 
other alligatorid crocodilians to such a degree as to render detailed 
description unnecessary. 

Measurements 

Left ulna Left radius 

Length 78 67 

Maximum thickness of shaft 5 6 

Ratio, length over length of humerus . . 764 656 

Carpus and Manit^s 

The carpus and manus are well preserved on both sides. The 
carpus consists of a very large radiale and a very small ulnare. The 
radiale is usually considerably larger than the ulnare in crocodilians, 
but in this case the discrepancy is size between the two bones is un- 
usually great. In form these bones do not differ materially from those 
of living crocodilians. 

Four metacarpals (I- IV) are preserved in the left manus and four and 
a half in the right (I-IV, | V). Metacarpals II and III are subequal in 
size, and are much larger than I and IV. I is moderately short and 
stout. II is longer than I, being the longest of the five, and is more 
slender than I. 

The phalanges are well preserved. The formula appears to be 
2.3.4.3.2. The third digit is somewhat longer and stouter than the 
others. 

Pelvic Girdle 

Both ilia are preserved, the left one sufficiently well for comparison 
with the iUa of other forms. Its vertical diameter is relatively greater 
in proportion to its length than mAUigator inississippicnsis. The length 
of the inferior border, including the pubic and ischiadic peduncles, is 
relatively greater than in the Florida alligator. The anterior border 
of the bone is less oblique to the superior and inferior surfaces than 
in the latter species. 

On the inner surface the rugosities that connect with the sacral 
vertebrae are definitely separated by a short space of smooth bone. 
In A. mississippiensis the rugose areas are continuous. 



mook: osteology of alligator prenasalis 33 

Measurements and Ratios 

Mus. Comp. Am. Mus. 

Zool. 1,014 31822 

Length of ilium 62 mm. 97 mm. 

Height 43 58 

Ratio, height over length 693 598 

Length of inferior border 40 61 

Ratio, length of inferior border over length . 663 628 

The left ischium is fragmentary. The right ischium is fairly well 
preserved, but its pubic peduncle is missing. The bone is relatively 
short, and the pubic process is elevated considerably above the level 
of the iliac articular surface. The anterior border of the bone is bent 
sharply inward. No such character is present in the ischium of the 
Florida alligator. 

Measurements and Ratios 

Mus. Comp. Am. Mus. 
Zool. 1,014 31,822. 

Left ischium Left ischium 

A Length, iliac peduncle to tip 57 101 

B Length, total 64 est. 107 

C Breadth of distal end 32 54 

D Breadth of shaft 14 21 

D/A 245 208 

C/A 561 534 

Both pubes are present in the specimen, but they are poorly pre- 
served. So far as they show any characters at all they resemble the 
corresponding bones of A. mississippiensis. 



Femur 

Both femora of No. 1,014 are well preserved. They resemble the 
femur of the Florida alligator closely. The head is somewhat flatter 
than in the latter species, but that may be partly due to crushing. 
The apex of the deltoid crest is slightly nearer the center of the shaft 
than in the living form; the outline of the distal end is somewhat more 
irregular. 



34 



bulletin: museum of comparative zoology 



-^v^ 






1^ 

2 



Fig. 6. Alligator prenasalis (Loomis). A, left femur, (reversed) posterior 
view; B, right tibia, anterior view; C, right fibula, internal view; D, right pes, 
superior view. One-half natural size. All views Mus. Comp. Zool. No. 1,014, 
except ungual phalanx marked x, which is No. 1,015. 



Measurements and Ratio 



Length, median 

Length, maximum 

Breadth, proximal end 

Breadth, distal end 

Breadth of shaft, minimum 

Circumference of shaft, minimum .... 
Index: circumference over maximum length 
A Center of proximal end to center of 4th 

trochanter 114 

B Center of 4th trochanter to center of distal 

end 203 

Ratio A over B 561 

Ratio, length, median, over length of tibia . 1448 
Ratio, length, median, over length of fibula . 1495 



t< 


o 


.s 


■^ 


s 

c 
o 


o 

s 


n §: 


2 :2 


o -s 


O *^ 


<N « 


•3 S 


2 fe 


i 1 


00 .2 


'^ 1 


. u a 

M 3 „ 




2 3 5 


dSfe, 


Mu 
fern 
dilu 


ill 


9) O 


t. -^ S 


u ♦^ s> 


t- -^ to 


. -^ 5 


£j= § 


«J3.S 


v^:^ 


mJSU: 


l£5^ 


l£S 


^■B 




325mm. 


174 


183 


129 


317 


164 


179 


124 


68 


42 


44 


29 


71 


36 


45 


30 


30 


21est. 


21 


10 


120 


72est. 


78 


49 


369 


413est. 


435 


380 



63 



67 



45 



101 


112 


79 


623 


598 


580 


1333 


1251 


1227 


1366 


1306 


1278 



MOOK : OSTEOLOGY OF ALLIGATOR PRENASALIS 35 



Tibia and Fibula 

Both tibiae and the left fibula are preserved in No. 1,014 and the 
right fibula in No. 1,015. These bones are somewhat more slender 
than the corresponding bones of Alligator mississippiensis, but other- 
wise they are not especially characteristic. The tibiae are somewhat 
flattened at their distal ends, but that may be due to crushing. 







'«■ 








•■» 








60 




m 


'**< 


e 




tH 


i-H 


•S 




O^ 


o 


..^ 




'-'•2 


rH 






, 


N 


nI 


t5 


00-2 

ro.2 


00 

CO 


Comp. 
alor pr 
ibia 


d 


»S 


^ w 


Si2 

3 


Mu; 

a tor 
ibia 


3 « 

^1 


. Ol-^ 


,ca 


u. en** 


l.« 


«^ »j 


M »J 


o:;: ^ 


2-^ 


s~«« 


s«« 




12 


;55^ 


^^ 


116 


98 
13 
10 
23 
235 


144 
38 
36 
69 

469 


148 


25 


23 


22 


17 


35 


41 


;th 301 


277 



Length total 

Breadth proximal end . . 
Breadth distal end . . . 
Circumference, minimum . 
Index, Circumference over length 



Tarsus and Pes 

The left astragalus of No. 1,015 is the only bone from the tarsal 
series that is completely preserved. Its form is indistinguishable from 
that of the astragalus of Alligator mississippiensis. 

Maximum breadth 24 mm. 

A number of isolated bone nodules probably represent other tarsal 
or carpal bones incompletely preserved. 

The pes is incomplete in both specimens. A composite from both 
1,014 and 1,015 provides a right pes with the proximal phalanx of the 
first digit missing. All of this pes is No. 1,014 except the claw of Digit 
IV, which is No. 1,015. A composite left pes is less complete. A 
number of duplicate bones belonging to No. 1,015 have necessarily 
been omitted from the composite. The pes resembles that of A. 
mississippiensis very closely except that the whole structure is some- 
what narrower in proportion to its length. 



36 



bulletin: museum of comparative zoology 



Vertebrae j 

A number of vertebrae are preserved in both 1,014 and 1,015. 
They overlap sHghtly with respect to position in the column, but 
for the most part occupy different places in the column. Seventeen 




Fig. 7. Alligator jn-enasalis (Loomis). Vertebra, Dorsal 3; A, anterior 
view; B, lateral view, left side; C, posterior view. One-half natural size. Mus. 
Comp. Zool. No. 1,014. 




Fig. 8. Alligalor prenasalis (Loomis). Vertebrae (1), Cervical 8; (2), 
Dorsal 1; (3), Dorsals 2, 3; (4), Dorsal 4; (5), Dorsals 5-6. A, Anterior views; 
B, lateral views, left side; C, posterior views. One-half natural size. Mus. 
Comp. Zool. No. 1,015, (1, 3, 4, 5); and No. 1,014, (2). 



MOOK : OSTEOLOGY OF ALLIGATOR PRENASALIS 



37 



of the best preserved vertebrae have been selected for description and 
illustration. These range from Cervical to Caudal 4 as follows: 

Cervicals 4, 5, 6 No. 1,014 

Cervical 8 1,015 

Dorsal 1 1,014 

Dorsals 2,3 1,015 

Dorsal 3 1,014 

Dorsal 4 1,015 

Dorsals 5, 6 1,015 

Dorsal 10? 1,014 

Dorsal 11? 1,015 

Lumbar 2 1,015 

Lumbar 3 1,015 

Caudal 1 1,015 

Caudal 4 1,014 




B 



•*<- 






. — — -rri^ r> 



Fig. 9. Alligator prenasalis (Loomis). Vertebrae, (1), Dorsal 10; (2), 
probably Dorsal 11; A, anterior views; B, lateral views, left side; C, posterior 
views. One-half natural size. Mus. Comp. Zool. No. 1,014 (1) and 1,015 (2). 



38 



bulletin: museum of compakative zoology 



Other vertebrae are present in the lot, but are not sufficiently dis- 
tinctive to warrant an attempt at accurate identification or description. 

These vertebrae are normal alligatoroid vertebrae, and differ from 
those of the living Florida alligator in no essential characters. 

Discrepancies in the measurements may be explained as due to 
differential distortion of the specimens. 









Fig. 10. Alligator prenasalis (Loomis). Vertebrae (1), Lumbar 2; (2), 
Caudal 1; (3), Caudal 4; A, anterior views; B, lateral views, left side; C, 
posterior views. One-half natural size. Mus. Comp. Zool. Nos. 1,014 (2) 
and 1,015 (1 and 3). 



mook: osteology of alligator prenasalis 



39 







Measurements 












4-9 


A 

i| 

pa a 


u a 

sa 

-St 
II 


1 

II 


Cervical 4.(1,014) 


24 est. 


44 


24 est. 


23 est. 


28 


Cervical 5.(1,014) 


23 est. 


44 


24 est. 


22 est. 


30 es.t 


Cervical 6.(1,014) 


28 


44 


25 est. 


— 


33 est. 


Cervical 8.(1,015) 


26 


55 est. 


20 


24 


— 


Dorsal 1.(1,014) 


27 


49 dist. 


20 est. 


26 est. 


49 


Dorsal 2.(1,015) 


25 


51 est. 


19 dist. 


25 


44+ 


Dorsal 3.(1,015) 


28 


52 


— 


28 


44+ 


Dorsal 3.(1,014) 


28 


48 est. 


28 


26 


58 


Dorsal 4.(1,015) 


27 


55 


26 dist. 


35 


— 


Dorsal 5.(1,015) 


27 


53 


27 


29 


— 


Dorsal 6.(1,015) 


29 


50 


— 


29 


54 est. 


Dorsal 10? (1,014) 


26 


48+est. 


33 


24 est. 


70+ 


Dorsal 11?(1,015) 


30 


50 est. 


33 


— 


— 


Lumbar 2.(1,015) 


28 


43 


33 


33 


76+ 


Lumbar 3.(1,015) 


30 


43 


32 est. 


34 est. 


70+ 


Caudal 1.(1,015) 


29 


43 est. 


— 


24 dist. 


— 


Caudal 4.(1,014) 


27 


— 


— 


20 


— 



Ribs 

A number of ribs are preserved, but they are not sufficiently dis- 
tinctive to warrant special description. 



Conclusions 

Alligator prenasalis was originally described by Prof. F. B. Loomis 
in the American Journal of Science as Crocodilus prenasalis} 

It was based upon fragments of the snout and lower jaws in the 
Museum of the South Dakota School of Mines. Ten vertebrae, twenty- 
five dermal scutes, a femur, a tibia, an astragalus, and some fragments 
in the Museum of Amherst College were treated as paratypes. The 
tj'pe and paratype specimens were found six miles apart, near Fulsom 
Post Office, South Dakota, in the Titanotherium Beds of the White 
River Formation. 

In 1916 Dr. M. G. MehF referred this species to his new genus 

1 Series 4, 18, 1904, pp. 427-432, 11 figs. 
^ Journ. Geol., 24, pp. 5.5, 56. 



40 bulletin: museum of comparative zoology 

Caimanoidra, and in 191S Dr. \Y. D. Matthew referred to it as Alli- 
gator "Crocodilus" prenasalis} 

This reference is correct, and the species stands as: Alligator preiia- 
salis (Loomis). 

The new specimens of the Museum of Comparative Zoology agree 
in characters with the type of .1. prenasalis in so far as comparison is 
possible, except that the type snout is somewhat less broad. As the 
type is a smaller and younger specimen this difference is quite evidently 
due to age differentiation. The level of the new specimens is practically 
identical with that of the type, and the locality is not far from the type 
locality. We need have no hesitation in referring the specimens under 
consideration to Alligator prenasalis. The new material is so much 
more complete than the type that it permits a more detailed descrip- 
tion of the osteology of the species than has heretofore been possible. 

The characters of the skull especially, show the species to be a true 
alligator, that is, a member of the genus Alligator Cu\aer. The region 
of the external narial aperture agrees with the type of Caimanoidea 
visheri of ]Mehl. The latter author separated Caimanoidea from 
Alligator partly on the characters of the narial aperture. In both 
species of Alligator now living the aperture is completely divided by a 
median septum. In Jacare and Caiman the aperture is not divided. 
In this respect Caimanoidea reseml)les Jacare and Caiman rather than 
Alligator. Alligator thomsoni, from the Snake Creek Beds, is inter- 
mediate in this respect between Caimanoidea visheri and the living 
alligators. It seems doubtful that this character is of generic value, 
and it is hardly sufficient ground, by itself, for separating Caimarioidea 
from Alligator. This does not affect the validity of Caimanoidea, as it 
is based on other characters as well. The material herein described 
also resembles the Paleocene and Eocene species that I have designated 
as Allognathosuchus, but its affinities are definitely closer to the living 
alligators than to the members of this genus. The genera Alligator 
Cuvier, Caimanoidea ]Mehl, and Allognathosuchus Mook, are closely 
related to each other. Jacare Gray and Caiman Spix exhibit resem- 
blances to these three genera, but they are evidently largely the 
results of parallelism rather than of close relationship. Brachychainpsa 
Gilmore, of the Cretaceous, may be ancestral to both the North 
American and South American phyla, as represented by Allognathosu- 
chus, Caimanoidea and Alligator on the one hand and Jacare and 
Caiman on the other. A provisional interpretation of these relation- 
ships is indicated on p. 41. 

' Amer. Mus. Journ. 18. pp. 505, 506. 



mook: osteology of alligator prenasalis 



41 



Recent 



Pleistocene 



Oi 









CaiiiKui 
Jacarr L ^ 

sclerovs niqer latirostris § ao ^ o 




Pliocene . Alligator thomsoni 



Paleocene 



Cretaceous 



Brachychampsa 



EXPLANATION OF PLATES 



PLATE 1 



MooK— Osteology of Alligator Prenasali 



PLATE 1 



Alligator prcnasalis (Loomis). Skull and jaws, superior view. One-half natural 
size. Mus. Comp. Zo5l. No. 1,014. 



BULL. MUS. COMP. ZOOL. 



MooK. Alligator Osteology. Plate 1 




.r** 




•^:-^i-5 










¥■ 




HELIOTYPE CO. BOSTON 



PLATE 2 



MooK. — Osteology of Alligator Prenasalis 



PLATE 2 

Alligator prenasalis (Loomis). Skull and jaws, palatal view. One-half natural 
size. Mus. Comp. Zool. No. 1,014. 



BULL. MUS. COMP. 200L. 



MooK. Alligator Osteology. Plate 2 









HELIOTYPE CO. BOSTON 



PLATE 3 



MooK. — Osteology of Alligator Pienasalis 



PLATE 3 

Alligator -prenasalis (Loomis). Skull and jaws, lateral view, left side. One-half 
natural size. Mus. Comp. Zool. Xo. 1,014. 



BULL. MUS. COMP. 200L. 



MooK. Alligator Osteology. Plate 3 



•A 




•..i 



'^ 



HELIOTYPt CO. BOSTON 



Sgp 281932 M- 



^/7^ 



Bulletin of the Museum of Comparative Zoology 

AT HARVARD COLLEGE 
Vol. LXXIV, No. 3 



ON CERTAIN SIMILARITIES BETWEEN 
SLOTHS AND SLOW LEMURS 



By William L. Stkaus, Jr. and George B. Wislocki ' 



[From the Departments of Anatomy, Johns Hopkins University 

and Harvard University] 



CAMBRIDGE, MASS., U. S. A.: 

PRINTED FOR THE MUSEUM 

September, 1932 



PUBLICATIONS 

OF THE 

MUSEUM OF COMPARATIVE ZOOLOGY 
AT HARVARD COLLEGE 



There have been pubhshed of the Bulletin Vols. I to LXV, LXVII- 
LXXIV, of the Memoirs Vols. I to LI. 

The Bulletin and Memoirs are devoted to the publication of 
original work by the Officers of the Museum, of investigations carried 
on by students and others in the different Laboratories of Natural 
History, and of work by specialists based upon the Museum Collec- 
tions and Exploration. 

These publications are issued in numbers at irregular intervals. 
Each number of the Bulletin and of the Memoirs may be sold sepa- 
rately. A price hst of the publications of the Museum will be sent on 
apphcation to the Director of the Museum of Comparative Zoology, 
Cambridge, Massachusetts. 



Bulletin of the Museum of Comparative Zoology 

AT HARVARD COLLEGE 

Vol. LXXIV, No. 3 



ON CERTAIN SIMILARITIES BETWEEN 
SLOTHS AND SLOW LEMURS 



By W^illiam L. Straus, Jr. and George B. W'islocki 



[From the Departments of Anatomy, Johns Hopkins University 

and Harvard University] 



CAMBRIDGE, MASS., U. S. A.: 

PRINTED FOR THE MUSEUM 

September, 1932 



No. 3. — On Certain Similarities between Sloths and Slow Lemurs 

By William L. Straus, Jr. and George B. Wislocki 

[From the Departments of Anatomy, Johns Hopkins University 
and Harvard University] 

Introduction 

In a previous paper the present authors dealt with the blood vascu- 
lar plexuses found in the extremities of certain mammals (Wislocki 
and Straus, 1932). During our investigations we emphasized that the 
particular type of plexus or rete mirabile which we designated " blood 
vascular bundle" was peculiar to certain of the edentates (sloths, 
didactyl and tetradactyl ant-eaters, pangolin) and to the lorisine or 
slow lemurs {Nycticebus, Loris, Perodicticus). We interpreted the 
vascular bundle type of plexus in the extremities as an example of 
structural convergence in two remotely related groups of mammals. 
Common to these groups also appeared to be a certain resemblance in 
their modes of living: namely arboreal life, sluggishness and a charac- 
teristic use of the extremities for the maintenance of clinging postures. 
It seemed of interest to investigate the general problem of whether 
these two groups of animals possess other structural peculiarities 
which might be interpreted as convergences relating to their habitual 
postures and modes of progression. These further investigations have 
been concerned with the proportions and certain skeletal characters of 
trunk and limbs, because it is in these relationships especially that one 
might expect to find the most pronounced expression of convergent 
adaptations in the groups of animals under consideration. Briefly it 
may be stated that we believe we have found data indicative of con- 
vergences between lorisine lemurs and the arboreal Xenarthra. These 
data we shall present and discuss. 

Material 

For the present investigation we have a large series of individuals 
and genera of both Xenarthra and Prosimiae at our disposal, as is 
shown in Tables 1 to 7. The data used herein are based, as indicated, 
not only upon new and unpublished observations, but also in part 
upon the literature. 

We are indebted to IVIr. Gerrit S. Miller, Jr. of the U. S. National 
Museum, Dr. T. Wingate Todd of Western Reserve University, and 



46 bulletin: museum of comparative zoology 

Dr. William K. Gregory and Mr. H. E. Anthony of the American 
Museum of Natural History for the opportunity of studying specimens. 
We likewise wish to thank our colleague, Dr. Adolph H. Schultz, who 
has generously permitted us to make use of some of his unpublished 
measurements of the limbs and trunk of prosimians and sloths. 

Observations and Discussion 

Proportions of the extremities. One of the most striking characteris- 
tics of the sloths is the unusual length of the upper extremity, which 
surpasses that of the lower extremity.^ As can be seen in Table 1, 
this disproportion is not nearly so marked in the two-toed sloth 
{Cholocpus) as in the three-toed sloth {Brady pus). It will also be 
noted that in none of the other Xenarthra, either arboreal or terrestrial, 
is the fore-limb as long as the hind-limb. 

From the conditions in such ground-living members of the order as 
the armadillos, Busy pus and Cabassous, we may conclude that a rela- 
tively short fore-limb is probably the primitive edentate condition. 
Thus the relatively great length of the upper extremity in Bradypus 
appears as a specialization, an adaptation to its sluggish, clinging ar- 
boreal mode of life. Cholocpus has developed in the same direction, 
but without such extreme specialization. 

An explanation of the relatively shorter upper extremities of the two 
ant-eaters, both extremely arboreal, may perhaps be found in their 
possession of long, prehensile tails. Both of the sloths, on the other 
hand, have extremely reduced tails, of no possible use for prehension. 
Cyclopes and Tamandua have thus adapted themselves to arboreal life 
in a different manner than have the sloths. This adaptation of the 
didactyl and tetradactyl ant-eaters has involved the development of 
long tails for grasping, whereas in the sloths the tail has become vestig- 
ial, and the upper extremity greatly lengthened.^ In relative length 
of upper extremity Tamandua more closely approaches Cholocpus than 
does Cyclopes. 

Table 1 also reveals the fact that the Lorisinae, the group of lemurs 
including the genera Nycticcbus, Loris and Perodicticus , whose mode of 

• In this study, upper extremity (or fore-limb) has been regarded as consisting of humerus and 
radius alone, lower extremity (or hind-limb) of femur and tibia. Lengths of hand and foot have 
not been considered, since inclusion of these differently specialized distal segments in comparing 
the limbs of members of such widely separated orders as Edentata and Primates might lead to 
erroneous conclusions. 

2 Cyclopes possesses 40 or 41 caudal vertebrae, Tamandua 35 or 37 (Flower, Giebel). Brady- 
pus bias on the average only 10 (8-12) caudal vertebrae, Choloepus but 5 (4-6) (from specimens 
listed by Bateson, Flower and Giebel, and from our own observations on 2 skeletons of Bradypus). 
In other Xenarthra the conditions are intermediate, the number of caudal vertebrae varying 
from II to 12 in Tolypeules to 29 in Myrmecophaga and 31 in a Dasypus rwvemcincius (Flower, 
Giebel). 



STRAUS AND WISLOCKI: SLOTHS AND LEMURS 



47 



life approximates that of the sloths (Wislocki and Straus, 1932), ex- 
hibit specializations similar to those of the Bradypodidae in respect to 
length of limbs. In all of the other lemurs, including the fossil Eocene 
genus Kotharctiis, humerus and radius together are less than three- 
quarters the length of combined femur and tibia. This we may accept 
as approximating the primitive lemurine condition. Only in the 



TABLE 1 

Number of 
specimens 



Intermembral index 
Humerus + radius 

X 100 

Femur + tibia 



Tarsius (Schultz) 

Tarsius 

Dauhentonia 

Microcehus 

Chirogale (Mollison) 

Galago (Schultz) 

Galago 

Lemur (Schultz) 

Lemur 

LejAdolemur 

Lichanotus 

Propiihecus (Mollison) 

Indris (Mollison) 

Perodicticus (Schultz) 

Perodicticus 

Nycticebus (Schultz) 

Loris (Mollison) 

Notharctus'\ 



7 54.2(50.4-57.1) 

1 57.9 

1 70.9 

1 70.9 

1 70.0 

7 66.4(58.9-72.5) 

2 59.0(53.1 and 64.8) 
2 72.2(71.3 and 73.1) 
2 71.4(71.2 and 71.6) 
1 62.9 

1 55.7 

2 64.0(63.0 and 65.0) 

3 62.0(61.0-63.0) 
1 88.3 

1 90.4 

3 92.3(90.4 94.5) 

7 90.0(82.0-94.0) 

1 61.9 



Cabassous 

Dasrjpus 

Myrmecophaga 

Tamandua 

Cyclopes 

Choloepus 

Bradypus (Schultz) 

Bradypus 



1 68.6 

1 64.6 

1 86.4 

1 90.8 

1 61.8 

3 112.2(107.4-117.1) 

3 175.0(171.5-178.4) 

2 172.3(171.3 and 173.3) 



Table 1. Intermembral indices of Prosimiae and Xenarthra. All measure- 
ments figuring in this table and Tables 2 and 3 have been taken according to 
the methods outlined by Schultz (1929). The indices taken from Schultz 
have been calculated by us from measui-ements hitherto unpublished. 



48 bulletin: museum of comparative zoology 

Lorisinae does the upper extremity approach the lower extremity in 
length. It thus becomes apparent that the relatively great length of 
the upper extremity in the lorisine lemurs is a definite specialization, 
probably an adaptation to their sluggish life, peculiar to them among 
the lemurs. Similarly, the reduced number of caudal vertebrae in the 
slow lemurs indicates that this adaptation has paralleled that of the 
sloths rather than that of the ant-eat ers.^ Xotharctus possessed a 
long tail (Gregory, 1920), and so do most of the existing Lemuroidea. 
Exceptions such as Iiidris are best regarded as parallelisms of the 
Lorisinae produced by unknown factors. 

The differences in the extremities of the various animals are most 
apparent in Table 2. Here the extremities are compared with the 

TABLE 2 

Number of Humerus + radius Femur + tibia 

specimens X 100 X 100 

anterior trunk height anterior trunk height 



Lemur (Schultz) 2 69.8(67.2 & 72.3) 96.6(94.2 & 98.9) 

Perodidicus (Schuhz) . . 1 75.7 85.7 

Nycticebus (Schultz) ... 3 77.9(72.5-81.9) 84.5(78.9-90.6) 



Cholocpns 1 103.9 96.8 

Brculypus 1 154.4 88.2 



Table 2. Upper and lower extremities in percentage of anterior trunk height 
in Prosimiae and Xenarthra. The measurements of the limbs of the speci- 
men of Bradypus were taken by Dr. A. H. Schultz, the trunk height by one 
of the present writers. 

trunk height. It is seen that two factors contribute to the differences 
between the intermembral indices of the sloths, i.e. the upper ex- 
tremity of Bradypus is not only relatively longer than that of Choloe- 
jnis, but the lower extremity is shorter. The same differences are 
noted when comparing Perodidicus and Nycticebus with Lemur. This 
analysis of limb length emphasizes the convergence between the sloths 
and the lorises. It indicates that in their adaptations to a clinging 
sluggish mode of life the limbs have undergone similar specialization 
in the two groups. 

' The not clearly separable genera, Nyclicebus. iMris and SI/>nops. when considered together, 
possess, on an average, 9 (5-12) caudal vertebrae: Perodidicus 14 (9-20). In other lemurs the 
number ranges from 9 or 11 in Indris to as many as 28 in individual specimens of Microcehus 
Lemur, Lepidolemur, Mvoiicebus and Propilhecus. Giebel reports as many as 32 in one Tarsius. 



STRAUS AND WISLOCKI: SLOTHS AND LEMURS 49 

The shape of the thorax. Next to the extremities, we might expect 
the trunk to exhibit evidences of adaptive convergence, especially in 
the thoracic region. In a typical quadruped, such as the dog, the chest 
is narrowand deep, the greatest dimension being in the ant.-post. diam- 
eter. This is the condition in most of the Lemuroidea, the chest index 
being below 100 (see Table 3). In the prosimians, however, the dis- 
crepancy between thoracic breadth and depth is not so marked as in 
such an animal as the dog. Indeed, in Tarsius, which is in no sense 
quadrupedal but which progresses by leaping in frog-like fashion and 
in an upright position, the breadth of the chest occasionally surpasses 



TABLE 3 

Chest index chest breadth 



Number of 
Specimens 



chest depth 



XlOO 



Canis (Jackson) 1 74.5 

Tarsius (Schultz) 7 90.9(83.9-102.0) 

Tarsius 1 91.2 

Galago (Schultz) 7 89.3(85.5-95.5) 

Lemur (Schultz) 2 94.2(91.7-96.6) 

Perodicticus (Schultz) 1 109.1 

Nydicebus (Schultz) 3 104.6(100.0-111.6) 



Choloepus 1 128.6 

Bradypus (Schultz) 2 157.7(155.6 and 159.7) 



Table 3. Chest indices of Canis, Prosimiae and Xenarthra. 

the depth. In both Nycticebv^ and Perodicticus the chest is relatively 
broader than in other lemurine forms. Herein they surpass even 
Tarsius, for their chest indices are always 100 or above, breadth of 
thorax normally exceeding the depth. Similarly, in both genera of the 
Bradypodidae the chest indices are very high. In these animals the 
thorax is indeed relatively much broader than in the slow lemurs. 

The experiments of Jackson (1907) upon dogs, and the ontogenetic 
studies by Schultz (1926) on Bradypus, suggest that the adult shape of 
the thorax (as expressed by the chest index) is influenced, at least to 
some degree, by the habitual posture of the animal. This means noth- 
ing more than that the resultant force of gravity upon the thoracic 
viscera is transmitted to and influences the development of the sur- 



50 



bulletin: museum of comparative zoology 



rounding bony cage. It is not surprising, therefore, to find that the 
chests of the sloths are relatively broader than those of typical quad- 
rupeds. It is likewise significant that the broadening of the thorax is 
more extreme in Brady pus, which spends more of its time in a perpendic- 
ular position than does Choloeims (Wislocki, 1928). Of the lemurs, 
only the lorises assume protracted clinging postures. Thus it is in- 
teresting that they exhibit a greater thoracic broadening than do the 

TABLE 4 
Number of Presacral Vertebrae 



Chlamydophorits . 

Priodontes 

Cabassous 

Tolypeutes 

Tatusia 

Dasypus 

Mynnecophaga . . 

Tamandua 

Cyclopes 

Bradypus 

Choloepus 



20 



21 



22 



23 



24 



25 



26 



27 



3 

28 



28 



1 
22 



29 



30 



1 



31 



32 



33 



3 9 



34 



Table 4. Distribution of specimens of Xenarthra with regard to the total 
number of presacral vertebrae. Based on data taken from Giebel (original 
observations), from Flower and from Bateson, and supplemented by the 
authors' own observations on the following specimens: 1 Cabassous, 2 Dasy- 
pus, 2 Cyclopes, 2 Bradypus, 2 Choloepus. The generic names given by Gie- 
bel (1874) and by Flower (1884) have been changed to agree with the nomen- 
clature used by Palmer (1904) and Miller (1924), and have been included 
in this table and in Table 6 in their revised forms. 



other Lemuroidea. Yet in this character, as in the relative lengths of 
the limbs, their specializations by no means appear to be as extreme as 
those of Brady pu^s and Choloepus. These observations lend support 
to the theory that habitual postures influence the shape of the thorax. 
They are also suggestive of an adaptive convergence between Lorisinae 
and Bradypodidae. 

The number of the presacral vertebrae. Another interesting feature 
concerns the number of the presacral vertebrae (see Tables 4 and 5). 



STRAUS AND WISLOCKI : SLOTHS AND LEMURS 



51 



In the terrestrial Xenarthra the number of presacral segments ranges 
from 20 to 25. This number shows a general increase in the arboreal 
members of the group: 24 or 25 in Cyclopes, 26 to 28 in Tamandua, 26 
to 29 in Brady pus, and 30 to 34 in Choloeptts. Since there is evidence 
that the primitive mammalian number of presacral vertebra was 26 
(Todd, 1922), it follows that in the recent terrestrial Xenarthra the 

TABLE 5 
Number of Presacral Vertebrae 



Tarsius 

Daubentonia . . . . 

Galago 

Chirogale 

Microcebus 

Lepidolemur . . . . 

Myoxicebus 

Lemur 

Ldchanotus 

Projrithecus .... 

Indris 

Perodidicus .... 
Nydicebus-Loris . 



25 26 27 28 29 29H 30 31 



2 

5 



14 

7 

12 



33 



2 
1 
2 



4 
3 
1 



3 
1 



1 
3 



1 



3 
21 



9 



Table 5. Distribution of specimens of Prosimiae with regard to the total 
number of presacral vertebrae. Based on data taken from the literature and 
from the authors' own observations of the following specimens: 3 Tarsius, 
8 Galago, 1 Chirogale, 1 Microcebus, 1 Lepidolemur, 1 Myoxicebus, 16 Lemur, 
3 Perodidicus, 8 Nydicehus-Loris. Information regarding the number of 
cervical vertebrae of some specimens was not available; but in view of the 
extreme numerical constancy of this series in primates, it has been deemed 
justifiable to assiune their number as being seven in all such instances. 



trunk has become shortened, the pelvis migrating cranially. The op- 
posite has occurred among some of the arboreal members of this group, 
at least to a marked degree in one representative, Choloepiis. Simi- 
larly, among the Lemuroidea, we can observe two trends in respect to 
the number of presacral vertebrae. Some lemurs have retained, on an 
average, the theoretical primitive number of 26. Myoxicrhus and oc- 
casional individuals of other genera have slightly shortened the pre- 
sacral column. Yet the tendency, in general, has been towards a 



52 



bulletin: museum of comparative zoology 



lengthening (see also Todd). This lengthening, or increase of pre- 
sacrals, is most marked in Xycticcbus, Loris and Pcrodicticus. The 
former are on the whole more advanced in this respect than is Pcrodic- 
ticus. This increase of presacral segments in the lorises and the ar- 
boreal Xenarthra may be interpreted as a convergence. 

The number of ribs. Coincident with its greater number of presacral 
vertebrae, Cholocpiis exhibits more pairs of ribs than do any of the 
other Xenarthra (see Table 6). Similarly, the arboreal Xenarthra 

TABLE 6 
Number of Pairs of Ribs 





9 


10 


11 


12 


13 


14 


15 


16 


17 


18 


19 


20 


21 


22 


23 


24 


Chlamydophorus 
PriodoJttes . 






1 


1 

2 


1 




2 


2 


2 


3 






1 


4 


7 




Cabassous 










Tolypeutes 






2 
5 
4 




Tatusia 

Dasypus 

Myrtnecophaga 
Tamandua 


2 

1 


3 


















Cyclopes 










. . . 




1 

31 


3 

1 




Bradypus 










1 


11 




Choloepus 










3 

































Table 6. Distribution of specimens of Xenarthra with regard to the number 
of pairs of ribs. Based on data taken from Giebel, from Flower and from 
Bateson, and from the authors' observations on the specimens Hsted in Table 
4 and on an additional skeleton of Myrmecophaga. 



possess more ribs than do the terrestrial, Myrmecophaga standing out 
as the sole exception to the rule. The same difference exists when the 
slow lemurs are contrasted with the other Lemuroidea, for the Lorisinae 
exhibit more ribs than do any of the remaining lemurs (see Table 7). 
The presence of a large number of ribs, as in Choloepus, could be re- 
garded as an adaptive specialization associated with the need for an 
effective armature for a slow-moving animal, were it not for the fact 
that Brady pus normally possesses no more ribs than does the more 
active terrestrial Myrmecophaga. Even when we consider that the 
eighth and ninth vertebral segments of Bradypus seem to have lost 
their ribs and become incorporated into the neck, the caudal extension 



STRAUS AND WISLOCKI: SLOTHS AND LEMURS 



53 



of the thorax by no means approaches that of ChoJocpus. Yet the 
tridactyl sloth is actually more sluggish than the didactyl sloth. Un- 
doubtedly other factors come into play in the production of a long 
thorax. 

The character of the processi spinosi. In addition to numerical verte- 
bral convergence, there are some interesting points which relate to the 
arrangements of the processi spinosi. It is well known that the 
.spinous processes of the vertebrae in typical active quadrupedal mam- 

TABLE 7 
Number of Pairs of Ribs 



Tarsius 

Daubentonia . . . . 

Galago 

Chirogale 

Microcebns 

Lepidolemur . . . . 

Myoxicebus 

Lemur 

Lichanotus 

Propithecus 

Indris 

Perodicticus 

Nycticehus-Loris . 



11 



12 13 



1 
3 



1 
1 

28 
3 
1 
1 



16 
4 

11 
3 



11 
1 



14 



1 

2 
1 



1 
5 



15 



3 
10 



16 



3 

14 



17 



18 



Table 7. Distribution of specimens of Prosimiae with regard to the number 
of pairs of rilss. Based on data taken from the Uterature and from the 
authors' own observations on the specimens Usted in Table 5. 



mals slope in two directions. The more cranial spines are retro verted 
and the more caudal anteverted toward an intermediate erect or anti- 
clinal spine. This anticlinal vertebra is situated near the caudal end 
of the thorax. This point is often referred to as the "center of 
motion." 

In all of the Xenarthra which we examined, both arboreal (Bradyjms, 
Choloepus) and terrestrial {Cahassous, Dasypus, Myrmccophaga), we 
have found no anticlinal spine. All of the presacral spines are directed 
back to the sacrum. In the slow lemurs, Nycticcbus and Perodicticus, 
the slope of the spinous processes is as in the Xenarthra. Other 



54 bulletin: museum of comparative zoology 

lemurs, such as Galago and Lemur, possess, on the contrary, an anti- 
cUnal spine, the process! spinosi being arranged in two sloping series. 

Wood Jones (1918) has dealt with this feature of the vertebral 
column at some length. According to him, uniform retroversion of the 
presacral spinous processes is apparently the primitive mammalian 
condition. The Bradypodidae, according to him, appear to \\axe been 
derived from a lumbering terrestrial stock, likewise possessing vuiiform 
spinous retroversion. The recent sloths, as well as the other Xen- 
arthra under consideration, appear therefore to have retained the an- 
cestral slope of the spinous processes. The same reasoning is not ap- 
plicable to the Lorisinae, because the ancestors of the slow lemurs were 
almost certainly active animals, in all likelihood possessing vertebral 
spines sloping in two series towards an intermediate anticlinal spine 
(also see Wood Jones). From what Gregory (1920) says of the simi- 
larity of the vertebral column of the Eocene genus Nothardus to that 
of the modern Lemur it is to be presumed that this fossil possessed an 
anticlinal spine. This merely bears out other evidence indicating a 
dual slope of the spines in the ancestors of Nydicchus and Pcrodiciicus. 
Thus in the lorises the uniform retroversion of the spinous processes, 
and the consequent absence of an anticlinal spine, appears to be a 
specialization, rather than a persistence of a primitive ancestral condi- 
tion. It may be interpreted as a convergence in the direction of the 
Bradypodidae. 

In the Bradypodidae the spines of the lower thoracic and of the 
lumbar vertebrae are characteristically low and Ijlunted, the cervical 
and upper thoracic spines longer and more pointed. Essentially' the 
same conditions obtain in the Lorisinae, but here the blunting is not so 
marked. This is illustrated by examining Nycticclms or Pcrodicticus. 
In Nycticehus the processi spinosi of all of the cervical and the upper 
thoracic vertebrae are rather long and well developed. Towards the 
lowermost part of the thorax the spinous processes become lower and 
blunter, a condition most pronounced in the proximal lumbar segments. 
Practically identical arrangements occur in Pcrodiciicus, but in this 
animal the cervical and uppermost thoracic spines are unusually long 
and pointed, much more so than in Xycticebus. In the more active 
members of their respective orders, such as Dasypus among the Xen- 
arthra, and Galago and Lemur among the Lemuroidea, an entirely 
difierent condition prevails. The thoracic and lumbar spines are 
prominent and relatively slender, while the cervical spines, excepting 
that of the epistropheus, are short and sometimes practically non- 
existent. It seems probable that here again in sloths and lorises we 



STRAUS AND WISLOCKI: SLOTHS AND LEMURS 55 

are dealing with the phenomenon of convergence. The relatively low, 
blunted thoracico-lumhar spines of the sloths and slow lemurs are pos- 
sibly correlated with back musculature more adapted to the gross 
purpose of strength and support than to the finer movements con- 
cerned with agility. The relatively greater development of the 
spinous processes of the cervical vertebrae may in turn be related to 
the development of the back musculature in this region. But natur- 
ally extended observations on the back musculature of these and re- 
lated types would be necessary to prove this assumption, which is 
offered here merely as a suggestion. 

Conclusions 

It is obvious that the sloths and lorises exhibit marked similarities 
in certain morphological characters. These similarities are apparently 
correlated with functional similarities of the parts involved. They are 
by no means indicative of any close genetic relationship of the two 
groups, but may be regarded as adaptive convergences. The sloths 
are plainly highly and peculiarly specialized members of the Xen- 
arthra, while the slow lemurs occupy a similar position among the 
Lemuroidea.^ Both groups have secondarily become adapted to a 
sluggish mode of arboreal life, and to the maintenance of clinging 
postures. Therein seems to lie the answer to the convergent structural 
adaptations. These structural similarities, which we interpret as 
convergences, involve the presence in the extremities of a peculiar and 
highly specialized type of blood vascular distribution (the bundle form 
of rete mirabile); increase in relative length of the upper extremity; 
reduction of the tail; relative broadening of the chest; increase in num- 
ber of presacral vertebrae and ribs; absence of an anticlinal vertebra ;2 
and blunting of thoraco-lumbar and lengthening of cervical spinous 
processes. 

' The two-toed and three-toed sloths differ markedly from one another, as is obvious from the 
characters under consideration (also see Wislocki, 1928). This is an interesting point, but it 
lies outside the scope of this paper. 

- Since this paper went to press one of us (W. L. S.) has had the opportunity of examining two 
skeletons of Loris lardigradus tardigradus recently acquired by the tj. S. National Museum. In 
each of these skeletons an anticlinal vertebra was clearly manifested. Loris thus appears to 
differ from both IVycticebus and Perodiclicus in respect to the slope of the vertebral spines, and in 
this character would seem to be more primitive than the other slow lemurs. It is possible, how- 
ever, that this particular feature of the vertebral column may be subject to individual variation, 
even within a single genus. 



56 bulletin: museum of comparative zoology 



BIBLIOGRAPHY 
Bateson, W. 

1894. Materials for the study of variation, treated with especial- regard 
to discontinuity in the origin of species. 598 pp., London and 
New York. 
Flower, W. H. 

1884. Catalogue of osteological specimens. Part II. Mammalia. 
Roy. Coll. Surgeons of England, 779 pp., London. 
GlEBEL, C. G. 

1874. Mammalia: Osteologie. Bronn's Klassen u. Ordnungen des Thier- 
Reichs, 6, pt. 5, p. 242. 
Gregory, W. K. 

1920. On the structure and relations of Nofharchis, an American Eocene 
primate. Mem. Amer. Mus. Nat. Hist., n. s., 3, p. 49. 
Jackson, C. M. 

1907. Is gravity the factor determining the thoracic index? Zeitschr. f. 
Morphol. u. Anthropol., 10, p. 240. 
Miller, G. S., Jr. 

1924. List of North American mammals (1923). Smithsonian Inst., 
U. S. Nat. Mus., Bull. no. 128. 

MOLLISON, T. 

1911. Die Korperproportionen der Primaten. Morphol. Jahrb., 42, p. 79. 
Palmer, T. S. 

1904. Index generum Mammalium. N. Amer. Fauna, no. 23, Biol. Sur- 
vey, U. S. Dept. Agric. 

SCHULTZ, A. H. 

1926. Fetal growth of man and other primates. Quart. Rev. Biol., 1, 

p. 465. 
1929. The technique of measuring the outer body of human fetuses and 
primates in general. Contrib. to Embryol., 20, no. 117, p. 213. 
Todd, T. W. 

1922. Numerical significance in the thoracico-lumbar vertebrae of the 
Mammalia. Anat. Rec, 24, p. 261. 
WiSLOCKI, G. B. 

1928. Observations on the gross and microscopic anatomy of the sloths 
(Bradypus griseus griseus Gray and Choloepus hoffmanni Peters). 
Jour. Morphol. and Physiol., 46, p. 217. 
WiSLOCKI, G. B. and W. L. Straus, Jr. 

1932. On the blood-vascular bundles in the limbs of certain edentates 
and lemurs. Bull. Mus. Comp. Zool., 74, no. 1, April, 1932, pp. 
1-16. 
Wood-Jones, F. 

1918. Arboreal man. 230 pp., London. 



DEC 1 4 1932 



Bulletin of the Museum of Comparative Zoology 

AT HARVARD COLLEGE 

Vol. LXXIV, No. 4 



FOSSIL TYPES OF FISHES, AMPHIBIANS, 

REPTILES AND BIRDS IN THE 
MUSEUM OF COMPARATIVE ZOOLOGY 



By \V. E. Schevill 



CAMBRIDGE, MASS., U. S. A.: 
PRINTED FOR THE MUSEUM 

December, 1932 

I" 



PUBLICATIONS 

OF THE 

MUSEUM OF COMPARATIVE ZOOLOGY 
AT HARVARD COLLEGE 



There have been published of the Bulletin Vols. I to LXV, LXVII- 
LXXIV, of the Memoirs Vols. I to LL 

The Bulletin and Memoirs are devoted to the publication of 
original work by the Officers of the Museum, of investigations carried 
on by students and others in the different Laboratories of Natural 
History, and of work by specialists based upon the Museum Collec- 
tions and Exploration. 

These publications are issued in numbers at irregular intervals. 
Each number of the Bulletin and of the Memoirs may be sold sepa- 
rately. A price list of the publications of the Museum will be sent on 
application to the Director of the INIuseum of Comparative Zoology, 
Cambridge, Massachusetts. 



Bulletin of the Museum of Comparative Zoology 

AT HARVARD COLLEGE 

Vol. LXXIV, No. 4 



FOSSIL TYPES OF FISHES, AMPHIBIANS, 

REPTILES AND BIRDS IN THE 
MUSEUM OF COMPARATR^E ZOOLOGY 



By W. E. Schevill 



CAMBRIDGE, MASS., U. S. A.: 

PRINTED FOR THE MUSEUM 

December, 1932 



No. 4. — Fossil Types of Fishes, Aviphibians, Reptiles and Birds in the 
Museum of Cmnparative Zoology 

By W. E. Schevill 



CONTENTS 

Page 

Introduction 59 

Fishes 60 

Reptiles and Amphibians 95 

Birds 99 



Introduction 

Ordinarily catalogues of this sort make mention of only primary 
types, as supplementary types are generally considered to be of slight 
importance. In paleontology it frequently happens that primary 
type material is poorly preserved or fragmentary. Subsequent col- 
lections may include better examples and may illustrate the species 
more successfully than the original specimens. A case in point is 
that of Cymhospondylus petrinus Leidy, which was described from 
three fragmentary vertebrae. Later Merriam published extremely 
detailed accounts of this species, based on practically complete skele- 
tons. These are mere plesiotypes, but are far more instructive than 
the holotype. On account of instances of this nature, all recognizably 
mentioned or figured specimens are included in this catalogue. Cf. 
Cushman, Proc. Boston Soc. Nat. Hist., 33, No. 6, 1907, p. 249. 

The accepted terminology of the primary types has been followed 
(holotjv-pe, cotype, paratype, etc., with the prefix "geno-" to indicate 
original material of a type species) ; for supplementary types the single 
term "plesiotype" has been used. Many have objected to this term 
as being too general to be of much definite use, but, as Schuchert says 
(Bull. U. S. Nat. Mus., 53, Pt. 1, 1905, p. 12), one "fails to see what 
can be gained by dividing these supplementary types into dilferent 
categories. After all, it is the primary types on which the taxonomy 
rests, and the supplementary types (plesiotvT)es) simply help to eluci- 
date the original material." Accordingly, such an elaborate system as 
that recommended by the Geological Society of America's Special 
Committee on the Marking of Type Specimens (Bull. Geol. Soc. Amer., 




60 • bulletin: museum of comparative zoology 

40, No. 1, 1929, pp. 215-220) has not been adopted; there, for ex- 
ample, are enumerated (op. cit., p. 219) five different kinds of supple- 
mentary (pubHshed) types, all of which are included in "plesiotype" 
as used in this catalogue. 

The types are listed alphabetically by genera, as far as the published 
references will permit, and are grouped under three headings: Fishes, 
Reptiles and Amphibians, and Birds. To save space by avoiding a 
certain amount of repetition, references to the literature are concen- 
trated in a bibliography at the end of the catalogue, arranged chrono- 
logically under the authors, who appear in alphabetical order. An 
exception is made for Agassiz's Poissons Fossiles, on account of the 
numerous variously dated fascicles in which the work was issued; each 
reference to this is given in full. 

In addition to the specimens listed on the following pages, the 
museum contains many of the selachian teeth described by St. John 
and Worthen in volumes 6 and 7 (1875 and 1883) of the Palaeontology 
of Illinois. These are for the most part in the Wachsmuth Collection, 
although some were obtained through Orestes St. John, and a few from 
other sources. They are not listed individually because of the dif- 
ficulty in identifying the actual types, when not especially labeled by 
the authors, from among large series of similar, specifically identical 
teeth; at best one can be certain, in such cases, only of the particular 
lot in which is the type. 

FISHES 

AcANTHASPis ARMATA Ncwbcrry 

5097. Plesiotj-pe. Columbus limestone. Sandusky, Ohio. Coll. 
H. Hcrzer, 1898. Eastman, 1907b, p. 117, pi. II, fig. 2; 1908, p. 145, 
pi. I, fig. 14. 

This individual is now Acanthaspis neivberryi Heintz. Holotype. 
Heintz, 1929/, p. 72, fig. 35. 

AcANTHOLEPis FRAGiLis Newbcrry 

5101. Plesiotype. Eastman, 1908, pp. 141-142, pi. Ill, figs. 5, 5a. 
See Eczematolepis fragilis (Newberry). 

Acanthopterygian, gen. et sp. ind. 

5067. Nilsson, 1824. p. 103, pi. II, fig. 1. 
See Semionotus nilssoni Agassiz. 



schevill: fossil types 61 

AcANTHURUs ovALis Agassiz 

5264 (Orig. No. 10). Plesiotype. Eocene. Monte Bolca, near 
Verona, Itaiy. Coll. Canossa, through Krantz, 1903. Volta, 1796, 
pp. 270-271, pi. LXV, fig. 1, as Chaetodon canus Seba, crrore; Agassiz, 
1835, lS35a, and Poiss. Foss., 4, p. 16* (1838), p. 253 (1842), as Fygaeus 
nohilis Agassiz, errore (Agassiz never saw the specimen, and made his 
identification from Volta's poor figure). 

Anaclitacanthus semicostatus St. John and Worthen 

5187. Genoholot^-pe. St. John and Worthen, 1875, p. 443. pi. 
XVI, figs. 14a-c. 

See Cten acanthus semicostatus (St. John and Worthen). 

Apedodus priscus Leidy 

5116, 5117. 2 plesiotypes. Chemung. Warren, Pa. Eastman, 
1907b, p. 166, pi. I, figs. 1, 2. 

Arthrodire, gen. et sp. ind. 

5237 Plesiotype. Baseof Waverly. Boyle County, Ky. Moritz 
Fischer leg., 1907. Dorsomedian. Eastman, 1908, fig. 32, p. 206. 

ASTEROLEPIS sp. 

5079. Hussakof, 1906, p. 130, fig. 13c. 
See AsTEROLEPis maxima Agassiz. 

AsTEROLEPIS MAXIMA AgaSsiz 

5079. Plesiotype. Upper Old Red. Nairnshire, Scotland. East- 
man, 1904b, p. 256, fig. 3, as Bothriolejns major (Agassiz); Hussakof, 
1906, p. 130, fig. 13c, as Asterolepis sj).; Eastman, 1907b, pp. 53-54. 

5081. Plesiotype. Upper Old Red. Nairnshire, Scotland. East- 
man, 1904b, p. 255, fig. 2, as Bothriolcpis major (Agassiz); 1907b, pp. 
53-54. 

Belemnacanthus giganteus Eastman 

5080 (formerly M.C.Z. 3091). Genoholotype. Mid-Devonian. 
Kerpen, Eifel, Germany. Coll. Schultze, 1871. Eastman, 1898, pp. 
552-556, fig. 50, p. 553.*^ 



62 bulletin: museum of comparative zoology 



Beloxorhynchus dayi Raymond 

1564. Holotype. Spray River shale. Massive Siding, Alberta. 
Joseph P. Day. Jr leg. and don., 1924. Raymond, 1925, pp. 551-554, 
fig. 1, p. 552 (text refers to 1654, error e). 

BiRKENiA ELEGANS Traquair 

1569, 1573. Two plesiotypes. Downtonian. Seggholm. A\Tshire, 
Scotland. D. Tait and H. C. Stetson leg., 1926. Stetson, 1928a. 
fig. 5, p. 465, and fig. 2, p. 461. 

2006, 2007. Plesiotypes. Downtonian. Seggholm, x\\Tshire. H. M. 
Geol. Surv. leg., 1927. Stetson, 1928a, fig. 1, p. 46o/fig. 3, p. 463, 
and fig. 6, p. 467. 

BoTHRioLEPis CANADENSIS Wliiteaves 

,5'Jl^l ■■5HB+ Plesiotype Devonian. Seaumenac, Quebec. Stetson, 
1930, p. 29, pi. VI, fig. 1. 

BOTHRIOLEPIS MAJOR (Agassiz) 

5079, 5081. Plesiot^-pes. Eastman, 1904b. fig. 3, p. 256, fig. 2, 
p. 255. 

See AsTEROLEPis maxima Agassiz. 

Campodus VARIABILIS (Ncwberrv' and Worthen) 

749. Plesiotype. Pennsylvanian. Osage County, Kans. Coll. G. 
C. Merrill. Eastman, 1902b, pp. 59 ft", fig. 2, p. 64, pi II, pi. Ill, fig. 1. 

Campyloprion annectans Eastman 

2039. Genoholotype. Eastman, 1902a, p. 151, fig. 3, p. 152, pi. 
VIII, fig. 2; also p. 332. 

See Helicoprion annectans (Eastman). 

Caranx primaevus Eastman 

5070. Holotype. Eocene. Monte Bolca, near Verona, Italy. 
Coll. Krantz, 1903. Eastman, 1904, pp. 28-29, fig. B, pi. I, fig. 4. 

Carcharias (Prionodon) sp. ind. 

, 5146, 5147, 5148. Four teeth. Pacific red clay. Albatross 

I'^'A Station 3681. Eastman, 1903a, p. 188 pi. I, figs. 1-4."^ 



schevill: fossil types 63 

5149. Two teeth. Pacific red clay. Albatross Station 3683. 
Eastman, 1903a, p. 188, pi. I, figs. 5, 6. 

Carcharias megalodon Agassiz 

5074. Cotype. Agassiz, Poiss. Foss., 3, pi. XXIX, figs. 6, 6a 
(1835). 

See Carcharodon megalodon Agassiz. 

Carcharias sulcidens Agassiz 

5075. Two cotypes. Agassiz, Poiss, Foss., 3, pi. XXXa, figs. 5, 

5\ 6, 61 (183/). /^ 

See Carcharodon rondelet/ii Muller and Henle. //^ 

Carcharodon Lu^nciformis Gibbes 

5163. Three plesiot^-pes. Pacific dark brown clay. Albatross 
Station 4685. Eastman, 1906a, pp. 80, 82, pi. II, figs. 13, 19, 22. 

5165. Plesiotype. Pacific fine green mud. Albatross Station 
4656. Eastman^ 1906a, pp. 80, 82, pi. II. fig. 21. 

5166. Plesiotype. Pacific globigerina ooze. Albatross Station 
4732. Eastman^ 1906a, pp. 81, 82, pi. II, fig. 20. 

Carcharodon megalodon Agassiz 

5074. Cotype. Miocene? St\Tia. Coll. Bronn. 1859. Agassiz, 
Poiss. Foss., 3, p. 249 (1843), pi. XXIX, figs. 6, 6a (1835). Referred 
to Carcharias on the plate. 

5144. Three plesiotypes. Pacific red clay. Albatross Station 
3691. Eastman, 1903a, p. 187, pi. II, figs. 31-33. 

5157. Three plesiotypes. Pacific red clay. Albatross Station 
3681. Eastman, 1903a, p. 186, pi. I, figs. 21-23. 

5167. Plesiotype. Pacific globigerina ooze. Albatross Station 
4740. Eastman]! 1906a, p. 82, pi. II. fig. 23. 

5265. Plesiot\T)e. Tertiary. Arica, Chile. A. E. Douglass leg. 
and don. Eastman, 1903a, p. 187. 

Carcharodon rondeletii Muller and Henle 

5075. Two plesiotypes (cotypes of C. sulcidens Agassiz). Pliocene. 
Castel - x\rquato, Italy. Agassiz, Poiss. Foss., 3, p. 254 (1843), pi. 
XXXa, figs. 5, 5\ 6, 6^ (1838), as C. sulcidens Agassiz (Carcharias in 
legend on plate). 



64 bulletin: museum of comparative zoology 

Carcharodon sulcidens Agassiz 

5075. Two cot>T)es. Agassiz, Poiss. Foss., 3, p. 254 (1843). 
See Carcharodon rondeletii Miiller and Henle. 

Catopterus gracilis J. H. Redfield 

5068 (formerly M.C.Z. 2531). Plesiotype. Eastman, 1905, pi. 
XIII, errore. 

See Dictyopyge macrura (W. C. Redfield). 

Chaetodon canus Seba "^^ 

5264. Plesiotype. Volta, 1796, pp. 270-271, pi. LXV, fig. 1, 
errore. 

See AcANTHURUS ovalis Agassiz. 

Chimaeroid (?) dermal plate 

5119. Eastman, 1908, p. 149, pi. 11, fig. 15. 
See under E]a,smobraneh. 

Chimaeroid (?) dermal tubercle 

5118. Kinderhook limestone. Burlington, Iowa. Coll. O. H. St. 
John, 1873. Eastman, 1908, p. 149, pi. II, figs. 12, 12a. 

Chomatodus inconstans St. John and Worthen 

5198. Plesiotype. Keokuk limestone. Keokuk, Iowa. Coll. A. H, 
Worthen, 1896. ' Eastman, 1903b, p. 204. 

Cladodus urbs-ludovici Eastman 

5179. Holotype. Devonian (New Albany black shale). Vicinity 
of Louisville, Ky. W. N. Longworth leg. Exch. Yale Peabody 
Museum. Eastman, 1908, p. 110, pi. Ill, fig. 3. 

Cladoselache acanthopterygius Dean 

5252 (Orig. No. E2474). Plesiotype. Cleveland shale. Lindale, 
Ohio. \V. L. Bryant leg. Buftalo Museum of Science don., 1930. 
Headless fish. Hussakof and Bryant, 1919, p. 128, fig. 43, and pi. 
XLV (the illustrations are of the counterpart of our specimen). 



schevill: fossil types 65 

CladoseLu\che pachypterygius Dean 

5243. Holotype (counterpart). Waverly of Kentucky. American 
Museum of Natural History don. Caudal fin. Dean, 1909, p. 241, 
p. 222, fig. 17 (the figured counterpart is in the American Museum, No. 
7583). 

Clupea sp. 

2001. Von Meyer, 1851a, pp. 89, 93, pi. XIY, fig. 4. 

2002. Von Meyer, 1851a, p. 89, pi. XVI, fig. 13. 
See Clupea humilis von Meyer. 

Clupea gracilis von Meyer 

1581 (Orig. No. 1011; B.S.N.H. 3476). Cushman, 1907, p. 270, as 
holotype, errorc (von Meyer designated no specimens, 1848, pp. 781, 
783; cf. von Meyer, 1851a, p. 92). 

See Clupea humilis von Meyer. 

Clupea humilis von Meyer 

1581 (Orig. No. 1011; B.S.N.H. 3476). Cotype. Molasse. Un- 
terkirchberg, near Uhn, Germany. Coll. Eser, through Boston 
Society of Natural History, 1925. Von Meyer, 1851a, pp. 88, 92, pi. 
XVI, fig. 12; Cushman, 1907, p. 270, as holotype of C. gracilis von 
Meyer, and plesiotype of C humilis von Meyer, erroribus. 

1585 (Orig. No. 985; B.S.N.H. 3446). Cotv^e. Molasse. Unter- 
kirchberg. Coll. Eser, through Boston Society of Natural History, 
1925. Von Meyer, 1851a, pp. 90, 92, pi. XIV, fig. 5 (referred with 
doubt); Cushman, 1907, p. 271, as cotype of C. ventricosa von Meyer, 
errore. 

1582 (Orig. No. 980; B.S.N.H. 3420). Plesiotype (cotype of C. 
lanccolata von Meyer). Molasse. Unterkirchberg. Coll. Eser, 
through Boston Society of Natural History, 1925. Von Meyer, 1851a, 
p. 88, p. 93, pi. XIV, fig. 2, as C. Icmceolata von Meyer. 

1583, 1584 (Orig. Nos. 983, 984; B.S.N.H. 3444, 3445). Plesiotype 
(holotype of C. ventricosa von Meyer). Molasse. Unterkirchberg. 
Coll. Eser, through Boston Society of Natural History, 1925. Von 
]\Ieyer, 1851a, pp. 90, 93, pi. XIV, figs, la, lb, as C. ventricosa von 
Meyer; Cushman, 1907, p. 271, as two co types of C. ventricosa von 
Me\'er, errore. 

2001 (Orig. No. 981; B.S.N.H. 3442). Plesiotype. Molasse. Un- 
terkirchberg. Coll. Eser, through Boston Society of Natural History, 



66 bulletin: museum of comparative zoology 

1925. Von Meyer, 1851a, pp. 89, 93, pi. XIV, fig. 4, as Clupea sp., 
intermediate between C. humilis von Meyer and C. lanceolata von 
Meyer; Cushman, 1907, p. 271, as cotype as C. lanceolata von Meyer, 
errore. 

2002 (Orig. No. 1007; B.S.N.H. 3472). Plesiotype. Molasse. 
Unterkirchberg. Coll. Eser, through Boston Society of Natural His- 
tory, 1925. Von Meyer, 1851a, p. 89, pi. XVI, fig. 13, as Clujjea sp. 

Clupea lanceolata von Meyer 

1582. Cotype. Von Meyer, ISSla, pp. 88, 93, pi. XIV, fig. 2. 
2001 (B.S.N.H. 3442). Cushman, 1907, p. 271, as cotype, errore. 
See Clupea humilis von Meyer. 

Clupea ventricosa von Meyer 

1583, 1584 (B.S.N.H. 3444, 3445). Holotype. Von Meyer, 1851a, 
pp. 90, 93, pi. XIV, figs, la, lb; Cushman, 1907, p. 271, as two of tliree 
cotypes, errore. 

1585 (B.S.N.H. 3446). Cushman, 1907, p. 271, as cotype, errore.  
See Clupea humilis von Meyer. 

COCCOSTEID 

1381. Dean, 1901, p. 122, footnote ("cranium"). 
See DiNiCHTHYS pustulosus Eastman. 

CoccosTEus DECiPiENS Agassiz 

1403. Plesiotype. Lower Old Red. Edderton, near Tain, Ross, 
Scotland. Coll.^Schultze, 1871. Stetson, 1930, p. 28, pi. VI, fig. 2. 

1409. Plesiotype. Old Red. Orkney. Coll. Damon. Dean, 
1909, p. 285, fig! 62G, p. 283. 

2041. Plesiotype. Old Red. Sandwick, Pomona, Orkney. Dean, 
1909, p. 284, fig. 62A, p. 283. 

CoRAX AFFiNis Agassiz 

5170. Plesiotype (cotype of C. appcndiculaivs Agassiz). Upper 
Cretaceous. Maestricht. Coll. Bronn, 1859. Agassiz, Poiss. Foss., 
3, p. 227 (1843), as C. appendiculatus; pi. XXVI, fig. 3 (1835), as 
Galeus appendiculatus. 



schevill: fossil types 



67 



Cor AX APPENDicuLATUs Agassiz 

5170. Cotype. Agassiz, Poiss. Foss., 3, p. 227, pi. XXVI, fig. 3. 
See Cor AX affinis Agassiz. 

CoRAX PRisTODONTUs Agassiz 

5169. Genocotype. Upper Cretaceous. Maestricht. Coll. Bronn, 
1859. Agassiz, Pofss. Foss., 3, pp. 224 f. (1843), pi. XXVI, fig. 9 (1835). 
Referred to Galeus on the plate. 

CoTTus BREVis Agassiz 

1586, 1592 (B.S.N.H. 3491, 3492). 2 plesiotypes. Von Meyer, 
1851a, p. 107, pi. XVI, figs. 7, 9, errore. 
See Lepidocottus multipinnatus (von Meyer), 

CoTTus (?) coNicus von Meyer 

1586, 1592. Two cotypes. Von Meyer, 1851a, p. 109. 
See Lepidocottus Multipinnatus (von Meyer). 

CoTTUS (?) multipinnatus von Meyer 

1587. Holotype. Von Meyer, 1851a, p. 106, pi. XVII, fig. 1. 
See Lepidocottus multipinnatus (von Meyer). 

Cottus papyraceus Agassiz 

5262 (formerlv M.C.Z. 2911). Holotype. Agassiz, Poiss. Foss., 4, 
p. 187 (1839), pi. XXXII, fig. 1 (1839). 
See Lepidocottus papyraceus (Agassiz). 

Ctenacanthus sp. 

5189. Pennsylvanian. Carlinville, 111. Coll. A. H. Worthen, 1896. 
Eastman, 1902b. p. 76. 

Ctenacanthus coxianus St. John and Worthen 

5188. Plesiotype. Kinderhook. Iowa. Coll. A. H. Worthen, 1896. 
Eastman, 1902b, p. 87. 

Ctenacanthus gracillimus Newberry and Worthen 

5184. Plesiotype. St. Louis limestone. Missouri? Coll. W'ard, 
1898. Eastman,' 1902b, p. 86, text fig. 12. 



68 



bulletin: museum of comparative zoology 




Ctenacanthus hybodoides Egerton 

5206 (formerly M.C.Z. 4213). Plesiotype. Coal Measures. Glas- 
gow. J. M. Campbell leg. Coll. 1\ Stock, 1883. Scale. Stock, 
1883, p. 185, pi. VII, fig. 16 (species credited to Agassiz). 

Ctenacanthus longinodosus Eastman 

5182. ParatN-pe. Kinderhook limestone. Burlington, Iowa 
("North hill exposure"). Coll. C. Wachsmuth, 1872. Young in- 
dividual. Eastman, 1902b, p. 80, text fig. ^, p. 79. 

Ctenacanthus semicostatus (St. John and Worthen) 

5187. Holotype (genoholotype of Anacliiacanthus semicostatus St. 
J. and W.). Upper Burlington fish bed. Burlington, Iowa. Coll. C. 
AVachsmuth, 1872. St. John and Worthen, 1875, p. 443, pi. XVI, figs. 
14a-c, as Anaclitacanthus ; Eastman, 1902b, p. 89. 

Ctenacanthus solidus Eastman 

5185. Paratype. Kinderhook. Iowa? Coll. A. H. Worthen, 1896. 
Eastman, 1902b, p. 90. 

Ctenacanthus varians St. John and Worthen 

5186. Holotype. Kinderhook (upper fish bed). Flint River, 
near Burlington, Iowa. Coll. C. Wachsmuth, 1872. St. John and 
Worthen, 1875, p. 422, pi. XIV, figs. 2a-i; Eastman, 1902b, p. 89. 

Ctenacanthus venustus Eastman 

5183. Holotype. Kinderhook. Iowa? Coll. A. H. Worthen, 1896- 
Eastman, 1902b, pp. 81-83, text fig. 10, p. 83. 

Cyprinus coryphaenoides Bronn 

5069, 5175 (Orig. Xos. 4a-b, 13). Two cotypes. Bronn, 1830, pp. 
20, 22, 28, pi. I, fig. 1 (synthetograph). 
See Leptolepis coryphaenoides (Bronn). 

Cyprinus papyraceus Bronn 

5258, 5260 (formerly M.C.Z. 2908, and 2914, 2922). Two cotypes. 
Bronn, 1828, pp. 377-381, pi. Ill, fig. 9 (s\iithetograph). 



schevill: fossil types 69 

5259 (formerly M.C.Z. 2931). Cotype. Bronn, 1828, pp. 377-381. 
See Leuciscus papyraceus (Bronn). 

Cyprinus priscus von Meyer 

1589 (Orig. No. 989; B.S.N.H. 3450). Cotype. Molasse. Unter- 
kirchberg, near Ulm, Germany. Coll. Eser, through Boston Society 
of Natural History, 1925. Von Meyer, 1848, p. 782; 1851a, p. 95, pi. 
XV, fig. 1. 

1588 (Orig. No. 988; B.S.N.H. 3449). Cotype. Molasse. Unter- 
kirchberg. Coll. Eser, through Boston Society of Natural History, 
1925. Von Meyer, 1851, p. 80; 1851a, p. 96, pi. XV, fig. 2 (composite 
figure, pectoral fin from counterpart). 

1591 (Orig. No. 1015; B.S.N.H. 3480). Cotype. Molasse. Un- 
terkirchberg. Coll. Eser, through Boston Society of Natural History, 
1925. Fin spine. Von Meyer, 1848, p. 782; 1851a, p. 97, pi. XV, 
fig. 5. 

1590 (Orig. No. 990; B.S.N.H. 3451). Plesiotype. Molasse. Unter- 
kirchberg. Coll. Eser, through Boston Society of Natural History, 
1925. Von Meyer, 1856, p. 22, pi. I, fig. 1; Cushman, 1907, p. 271, 
as cotype, errore. 

Deltodus contortus (St. John and Worthen) 

5137. Holotype (genoholotype of Taeniodus contortus St. John and 
Worthen). Lower Carboniferous. Vise. Belgium. Coll. L. G. de 
Koninck, 1861. St. John and Worthen, 1883, p. 76, as Taeniodus; 
Eastman, 1903b, p. 202, pi. IV, fig. 37, 43. 

Deltodus occidentalis (Leidy) 

5134. Plesiotype. Keokuk limestone. Keokuk, Iowa. Coll. A. H. 
Worthen, 1896. "^ Eastman, 1903b, p. 200, pi. IV, fig. 38. 

Deltodus occidentalis latior St. John and Worthen 

5127. Plesiotype. Keokuk limestone. Keokuk, Iowa. Coll. A. H. 
Worthen, 1896. Eastman, 1903b, p. 200, pi. V, fig. 53. 

Deltodus spatulatus Newberry and Worthen 

5121. Plesiotype. Burlington limestone. Burlington, Iowa. Coll. 
C. Wachsmuth,'l872. Eastman, 1903b, p. 200, pi. V, fig. 55. 

5122. Plesiotype. Burlington limestone. Augusta, Iowa. Coll. 
Beebe, 1898. Eastman, 1903b, p. 200, pi. IV, fig^ 42. 



// 



70 bulletin: museum of comparative zoology 

5139. Plesiotype. Keokuk limestone. Keokuk, Iowa. Coll. A. H. 
Worthen, 1896. Eastman, 1903b, p. 199, pi. IV, fig. 41. 

Dictyopyge macrura (W. C. Redfield) 

5068 (formerly M.C.Z. 2531). Plesiotype. Triassic. Locality un- 
known. Eastman, 1905, pi. XIII, as Catopterus gracilis J. H. Red- 
field, errore; 1911, p. 56. 

DiNICHTHYID 

1475. Cleveland shale. Vicinity of Cleveland, Ohio. Coll. Wm. 
Clark, 1896. Postero-ventro-median. Eastman, 1897c, p. 26, pi. V, 
fig. 1; 1907b, p. 144, and 1908, p. 205, as Titanichthys. 

DiNICHTHYS sp. 

1299 (Orig. No. 15). Plesiotype. Cleveland shale. Lorain County, 
Ohio. Coll.J. Terrell, 1885. Antero-ventro- median. Eastman, 1897c, 
p. 26. This specimen cannot be found in the collections. 

1300. (Orig. No. 20). Cleveland shale. Lorain County, Ohio. 
Coll. J. Terrell, 1885. Postero-ventro-median. Eastman, 1896, p. 47, 
as Titanichthys sp.; 1897c, p. 24, pi. II, fig. 2. 

DiNICHTHYS bohemicus (Barrandc) 

1376. Plesiotype. Devonian (Ggi). Chotec, Bohemia. Coll. 
Schary, 1882. Median dorsal. Eastman, 1897c, pp. 37-38, pi. V, 
fig. 2. 

1377. Plesiotype. Devonian (Ggi). Svagerka, Bohemia. Coll. 
Scharv, 1882. Median dorsal. Eastman, 18^c, pp. 33, 37-38, pi. 
II, fig. 3. 

DiNICHTHYS eifeliensis Kayscr 

1374. Plesiotype. Devonian. Berndorf, Eifel, Germany. Coll. 
Schultze, 1871. Carinal process of median dorsal. Eastman, 1897c, 
pp. 33, 36-37, pi. Ill, fig. 3. 

1474. Plesiotype. Devonian. Eifel. Coll. Schultze, 1871. Right 
antero-ventro-lateral. Eastman, 1897c, pp. 36-37, pi. V, fig. 4. 

DiNICHTHYS intermedius Ncwbcrry 

1343. Plesiot\'pe. Cleveland shale. Lorain County, Ohio. Coll. 
J. Terrell, 1885. ^ Left mandible. Stetson, 1930, p. 27, pi. V, fig. 2. 
1380. Plesiotype. Cleveland shale. Vicinity of Lindale, Ohio. 



schevill: fossil types 71 

Coll. Wm. Clark, 1S96. Antero-ventro-median. Eastman, 1897c, p. 26, 
pi. Ill, fig. 1. 

1477. Plesiotype. Cleveland shale. Vicinity of Cleveland, Ohio. 
Coll. Wm. Clark, 1896. Skull and jaw. Claypole, 1892, pp. 199-207, 
figures; Eastman, 1897c, pp. 19-20, pi. I, fig. 1 (cranium only); Stetson, 
1930, p. 29 (and fig. 2, p. 30 — s>Tithetograph with M.C.Z. 5244). 

DiNiCHTHYS PELMENSis Eastman 

1375. Holotype. Devonian. Pelm, Eifel, Germany. Coll. Schultze, 
1871. Median dorsal. Eastman, 1897c, pp. 33, 36, pi. II, fig. 4. 

DiNiCHTHYS PUSTULOSUS Eastman 

1381. Holotype. Mid-Devonian (Hamilton). Hydraulic cement 
quarries near Milwaukee, Wis. Coll. F. H. Day, 1881. Left antero- 
dorso-lateral. Eastman, 1897c, p. 39, pi. Ill, fig. 4; Dean, 1901, p. 
122, footnote, as Coccosteid. 

1389. Plesiotype. Mid-Devonian (Hamilton). White Sulphur 
Springs, two miles east of Buffalo, Iowa. Coll. J. A. Udden, 1898. 
Dorso-median. Hussakof, 1906, p. 142, fig. 22D, p. 141. 

1390. Plesiotype. Mid-Devonian (Lower Cedar Valley lime- 
stone). Vicinity of Rock Island, 111. A. S. Tiffany leg., 1883. 
Exch. with J. A. Udden, 1898.". Cranium. Eastman, 1907b, fig. 25, 
p. 132, pi. XII; 1908, p. 194, fig. 28, p. 197, pi. IV. 

5076. Plesiotype. Mid-Devonian (Hamilton). Buffalo, Iowa. 
J. A. Udden leg. and don., 1899. Fragmentary gnathal. Eastman, 
1902c, p. 657. 

DiNiCHTHYS TERRELLi Ncwbcrry 

1301 (Orig. No. 29). Plesiotype. Cleveland shale. Lorain County, 
Ohio. Coll. J. Terrell, 1885. xVntero-ventro-median. Eastman, 
1896. p. 47, as Tiianichthys sp.; 1897c, p. 25, pi. II, figs. 5-6. 

1315 (Orig. No. 30). Plesiotype. Cleveland shale. Lorain County, 
Ohio. Coll. J. Terrell, 1885. Carinal process of median dorsal. 
Eastman, 1897c, p. 33, pi. Ill, fig. 2. 

1325 (Orig. No. 31). Plesiotype. Cleveland shale. Lorain County, 
Ohio. Coll. J. Terrell, 1885. Left postero-dorso-lateral. Eastman, 
1897c, p. 21. This specimen is missing at present. 

1338 (Orig. No. 6). Plesiot\'pe. Cleveland shale. Lorain County, 
Ohio. Coll. J. Terrell, 1885. ' Right suborbital. Stetson, 1930, pp. 
23-25, fig. 1, p. 23. 



72 bulletin: museum of comparative zoology 

1379. PlesiotNT^e. Cleveland shale. Lindale, Ohio. Prentis Clark 
j^ leg. Coll. Wm. Clark. 1896. Antero- and postero-dorso-laterals. 

fh Eastman, 18%c, p. 20, pi. II, fig. 1. 

5242. Plesiotype. Cleveland shale. Ohio. Fragment of man- 
dible. Stetson, 1930, p. 27, pi. V, fig. 1. 

5244. Plesiotype. Cleveland shale. Ohio. Coll. J. Terrell, 1885. 
Mounted specimen. Stetson, 1930, pp. 19-26, pis. I, II, III, IV, VI, 
fig. 3 (and fig. 2, p. 30 — synthetograph with M.C.Z. 1477). 

5245. Plesiotype. Cleveland shale. Ohio. Postero-dorso-lateral. 
Stetson, 1930, p. 27. 

Dinomylostoma beecheri Eastman 

The holotype is in the Yale Peabody Museum (No. 2850), not, as 
said by Hussakof (1913, p. 247, footnote), in the Museum of Compara- 
tive Zoology. 

DioDON erinaceus Agassiz 

5071. Plesiotype. P^ocene. Monte Bolca, near Verona, Italy 
Eastman, 1904, p. 34, fig. D (fig. "4" in the text). 



/ 



DiPLODUS sp. 

5133. Knight, 1899, pp. 366, 372, /74. 
See Phoebodus knightianus Eastman. 

Dipnoan dental plate, gen. et sp. ind. 

5098. Eastman, 1907b, pi. IV, fig. 15, p. 203. 

See Synthetodus calvini Eastman. 

Upper Devonian (state quarry beds). Johnson County, Iowa. 
Eastman, 1907b, pi. II, fig. 1. This specimen cannot be located in the 
collections. 

5110. Eastman, 1907b, pi. IV, fig. 16. 

See DiPTERUS digitatus Eastman. 

DiPTERUS sp. 

5092. Udden, 1899, p. 494, fig. 1. 
See DiPTERUS uddeni Eastman. 

5102. Palatal tooth. Eastman, 1907b, pi. IV, fig. 12; 1908. pi. II, 
fig. 17, p. 301. 

5096, 5105. Eastman, 1907b, pi. IV, figs. 7, 2. 



schevill: fossil types 73 

See DiPTERUS pectinatus Eastman. 
5107. Eastman, 1907b, pi. IV, fig. 8. 
See DiPTERUS digitatus Eastman. 

DiPTERUS CALViNi Eastman 

5093. Holotype. Mid-Devonian (Cedar Valley limestone). Fair- 
port, Iowa. J. A. Udden leg. and don. Mandibulaj tooth. East- 
man, 1900a, pp. 38-39, fig. 7; 1907b, p. 160, pi. IV, fig. 1; 1908, p. 219, 
pi. II, fig. 1. 

DiPTERUS cosTATUS Eastman 

5094. Holotype. Upper Devonian (state quarry beds). North 
Liberty, Johnson County, Iowa. J. A. Udden leg. Coll. Calvin. 
1897. Mandibular tooth. Eastman, 1900a, p. 39, fig. 4, p. 38; 1907b, 
p. 161, pi. IV, fig. 9; 1908, p. 220, pi. II, fig. 8. 

DiPTERUS DIGITATUS Eastman 

5107. Cotype. Upper Devonian (state quarry beds). North 
Liberty, Iowa. Palatal tooth. Eastman, 1907b, pi. IV, fig. 8, as 
Dipterus sp.; 1908, p. 221, pi. II, fig. 6, and pi. VII, fig. 20. 

5108. Four cotypes. State quarry beds. Johnson County, Iowa. 
Mandibular teeth.' Eastman, 1908, p. 221, pi. VII, figs. 16-19, 25. 

5109. Four cotypes. State quarry beds. Johnson County, Iowa. 
Palatal teeth. Eastman, 1908, p. 221, pi. VII, figs. 21-24. 

5110. CotN-pe. State quarry beds. North Liberty, Iowa. Man- 
dibular tooth. Eastman, 1907b, pi. IV, fig. 16, as Dipnoan dental 
plate; 1908, p. 221, pi. II, fig. 21, and pi. VII, fig. 25. 

DiPTERUS MORDAX Eastman 

5095. Two cotypes. L^pper Devonian (state quarry beds). 
North Liberty, Johnson County, Iowa. J. A. Udden leg. Coll. 
Calvin, 1897. Mandibular teeth. Eastman, 1900a, pp. 39-40, figs. 
6, 8, p. 38; 1907b, p. 161, pi. IV, figs. 5, .6; 1908, p. 220, pi. II, figs. 4, 5. 

5113. Five plesiotypes. State quarry beds. Johnson County, 
Iowa. Mandibular teeth. Eastman, 1908, p. 220, pi. VII, figs. 5-9. 

DiPTERUS MURCHisoNi Pander 

5103. Plesiotype. Devonian. Berndorf, Eifel, Germany. Coll. 
Schultze, 1871. Mandibular tooth. Eastman, 1900c, p. 177; 1908, 
p. 226, pi. II, fig. 10. 



74 bulletin: museum of comparative zoology 

DiPTERUS nelsoni Newbcrry 

5111. PlesioU-pe. Chemung. Warren, Pa. C. E. Beecher leg. 
and don. Mandibular tooth. Eastman, 1907b, p. 163, pi. IV, figs. 
13, 14; 1908, p. 223, pi. II, figs. 11, 11a. and pL VII, fig. 3. 

5112. Two plesiotjs-pes. Chemung. ^Yar^en County, Pa. Pala- 
tal and mandibular teeth. Eastman, 1908, p. 223, pi. VII, figs. 1, 4. 

Dipterus pectinatus Eastman 

5096. Cotype. Upper Devonian (state quarry beds). North 
Liberty, Johnson County, Iowa. Coll. Calvin, 1896. Palatal tooth, 
Eastman, 1907b, pi. IV, fig. 7, as Dipterus sp. ; 1908, p. 222, pi. II, fig. 7. 

5105. Cotype. State quarry beds. North Liberty, Iowa. ISIan- 
dibular tooth. Eastman, 1907b, pi. IV, fig. .2, as Dipterus sp.; 1908, 
p. 222, pi. II, fig. 2, and pi. VII, fig. 13. 

5106. Five cotypes. State quarry beds. Johnson County, Iowa. 
Palatal and mandibular teeth. Eastman, 1908, p. 223, pi. VII, figs. 
10-12, 14, 15. 

Dipterus uddeni Eastman 

5092. Holotype. Mid-Devonian (Cedar Valley limestone). Buf- 
falo, Iowa. J. A. Udden leg. and don., 1899. Mandibular tooth. 
Udden, 1899, p. 494, fig. 1, as Dipterm sp.; Eastman, 1900a, p. 37, fig. 
5, b\ p. 38; 1907b, p. 160, pi. IV. figs. 3, 4; 1908, p. 218, pi. II, figs. 3, 3a. 

Drepanaspis gemundenensis Schliiter 

5218. Plesiotype. Devonian. Gemiinden, Hunsriick, Rhineland. 
T. Barbour don.,' 1928. Stetson, 1931, p. 150, fig. 5. 

5219. Plesiotype. Devonian. Gemiinden. T. Barbour don., 1928. 
Stetson, 1931, p. 151, figs. 6, 7 (synthetograph with M.C.Z. 5238). 

5238. Plesiotype. Devonian. Gemiinden. T. Barbour don., 1928. 
Stetson, 1931, p.'l52, fig. 7 (synthetograph with M.C.Z. 5219). 

Eczematolepis fragilis (Newberry) 

5101. Plesiotype. Corniferous. Falls of the Ohio, Ind. Eastman, 
1908, pp. 141, 142, pi. Ill, figs. 5, 5a, as Acantholepis. 

Edestus heinrichsii Newberry and AVorthen 

5190. Plesiotype. Pennsylvanian. CarHnville, 111. Coll. A. H. 
Worthen, 1896. " Eastman, 1902b, p. 76, text fig. 7. 



schevill: fossil types 75 

Elasmobranch dermal plate 

5119. Kinderhook limestone. Burlington, Iowa. Coll. O. H. St. 
John, 1873. Eastman, 1903h, p. 220, fig. 16; 1908, p. 149, pi. II, figs. 
15, 15a, as Chimaeroid (this figure is ca. x 2, not xt, as noted in the 
legend). 

5131. Two specimens. Kinderhook. Burlington, Iowa. Coll. 
O. H. St. John, 1873. Eastman, 1903b, p. 220, pi. V, figs. 56, 56a, 57. 

5135. Kinderhook. Burlington, Iowa. Coll. O. H. St. John, 1873. 
Eastman, 1903b, p. 220, pi. V, fig. 50. 

Elonichthys brownii (Jackson) 

5083 (B.S.X.H. 7900; formerly M.C.Z. 1961). Holotype. Mis- 
sissippian. Albert ^Nlines, New Brunswick. Exch. Boston Society 
of Natural History, 1897. Jackson, 1851, p. 22, as Palaconiscus 
broiniii; 1852, p. 139; Lambe, 1909, p. 169, figs. 3, 4, p. 173; 1910, p. 22, 
pi. I, fig. 2, pi. IV, fig. 1, pi. V, figs. 2, 3, 5, 6. 

5085 (B.S.N.H. 7901; formerly M.C.Z. 1957). Plesiotype. Mis- 
sissippian. Albert Mines, New Brunswick. Exch. Boston Society of 
Natural History, 1897. Lambe, 1910, pi. I, fig. 5, pi. IV, fig. 4. 

5088 (B.S.N.H. 7903; formerly M.C.Z. 1953). Plesiotype. Mis- 
sissippian. Albert Mines, New Brunswick. Exch. Boston Society of 
Natural History, 1897. Jackson, 1851, p. 24, as Palaconiscus; Lambe, 
1910, pi. II, fig. 7. 

Elonichthys elegantulus Eastman 

5221. Holotype. Mississippian. Albert Mines, New Brunswick. 
Exch. Boston Society of Natural History, 1897. Eastman, 1908, p. 
274. 

Elonichthys perpennatus Eastman 

5104. Holotype. Pennsylvanian. Mazon Creek, Grundy County, 
111. C. E. Beecher leg. and don., 1899. Eastman, 1902d, p. 539 
fig. 4; 1903b, p. 190, pi. V, fig. 49. 

Erismacanthus barbatus Eastman 

5138. Holotype. Kinderhook limestone. Burlington, Iowa. Coll. 
C. Wachsmuth,' 1872. Eastman, 1903b, p. 211, pi. V, fig. 47. 



76 bulletin: museum of comparative zoology 



Erismacanthus formosus Eastman 

Holotype. Mississippian (St. Louis limestone). Vicinity of St. 
Louis, Mo. Coll. Hambach. Eastman, 1902c, p. 850, fig. 1; 
1903b, p. 212, fig. 13. This specimen has not been found. 

Erismacanthus maccoyanus St. John and Worthen 

5199. Plesiotype. Kinderhook limestone. Le Grand, Iowa. Exch. 
United States National Museum, 1899. Eastman, 1903b, p. 212. 

FissoDUS dextatus Eastman 

5130. Holotype. Pennsylvanian. Topeka, Kans. S. A. Miller 
leg., 1898. Eastman, 1903b, p. 175, pi. II, fig. 12. 

Galeocerdo denticulatus Agassiz 

5077. Holotype. Upper Cretaceous. Maestricht. Coll. Bronn, 
1859. Agassiz,' Poiss. Foss., 3, p. 233 (1843), pi. XXVI, fig. 1 (1835). 
Referred to Galeus on the plate. Agassiz, 1856, p. 274, asPrionodon? 

Galeus appendiculatus Agassiz 

5170. Cotype. Agassiz, Poiss. Foss., 3, pi. XXVI. fig. 3. 
See CoRAX affinis Agassiz. 

Galeus denticulatus Agassiz 

5077. Holotype. Agassiz, Poiss. Foss., 3, pi. XXVI, fig. 1. 
See Galeocerdo denticulatus Agassiz. 

Galeus pristodontus Agassiz 

5169. Cotype. Agassiz. Poiss. Foss., 3, pi. XXVI, fig. 9. 
See CoRAX pristodontus Agassiz. 

GoBius (?) sp. 

5211. Von Meyer, 1856, p. 27, pi. I, fig. 6. 
See Lepidocottus multipinnatus (von Meyer). 

GoBius (?) conicus 

1586, 1592. Two cotypes. Von Meyer, 1851, p. 80. 
See Lepidocottus multipinnatus (von Meyer). 



schevill: fossil types 77 

GoBius MULTiPiNNATUS von Meyer 

1587. Holotype. Von Meyer, 1848, p. 783. 
See Lepidocottus multipinnatus (von Meyer). 

Harpacanthus fimbriatus (Stock) 

5201 (formerly M.C.Z. 4259). Genoholotype. Carboniferous lime- 
stone. Gilmerton, near Edinburgh, Scotland. W. Tait Kinnear and 
W. Anderson leg. Coll. T. Stock, 1883. Stock, 1883, p. 177, pi. VII, 
figs. 1, la, as Tristychius; Traquair, 1886, pp. 493-496. 

Stock's fig. la is an erroneous reconstruction. 

Helicoprion annectans (Eastman) 

2039. Holotype (genoholotype of Campyloprion anncdans East- 
man). Carboniferous. Locality unknown. Exch. Tufts College. 
Eastman, 1902a, p. 151, fig. 3, p. 152, pi. VIII, fig. 2, also p. 332; 1902b, 
pp. 64 sqq., text figs. 3, 4, and pi. IV; 1903, p. 286, footnote, pi. XXI, 
fig. 1. All these references are to Campyloprion Eastman. 

Helodus incisus Eastman 

5123. Holotype. Mississippian. Salem, Ind. Cassiday leg. Coll. 
O. H. St. John,' 1871. Eastman, 1903b, p. 204, pi. V, figs. 54a-b. 

5124. Paratype. Mississippian. Salem, Ind. Cassiday leg. Coll. 
O. H. St. John, \871. Eastman, 1903b, p. 205. 

Hemipristis sp. ind. 

5150, 5151. Two teeth. Pacific red clay. Albatross Station 
3681. Eastman. 1903a, p. 188, text fig. 5, pi' I, figs. 7, 8. 

HiSTIONOTOPHORUS BASSANII (dc Zigno) 

5176 (formerly M.C.Z. 3075. Plesiotype. Eocene. Monte Boica, 
near Verona, Italy. Coll. C. R. Eastman, 1903. Eastman, 1904, p. 
32, pi. I, figs. 1, la. 

5177 (Orig. Nos. 14, 15; formerly M.C.Z. 3074). Plesiotype. 
Upper Eocene. Monte Bolca. Coll. Krantz, 1903. Eastman, 1904, 
p. 32, pi. I, fig. 2. 

5178. Plesiotype. Upper Eocene. Monte Bolca. Eastman, 1904, 
p. 32, pi. I, fig. 3'. 

Text figure C is a sjiithetograph based on these three specimens. 



78 bulletin: museum of comparative zoology 



HoMACANTHUS ACiNACiFORMis Eastman 

5115 (formerly M.C.Z. 1625). Holot\i)e. Chemung. ^Ya^^en, 
Pa. Coll. F. A. Randall, 1897. Eastman, 1903b, p. 218, pi. V, fig. 
58; 1907b, p. 75, pi. I, fig. 16; 1908, p. 151, pi. Ill, fig. 10. 

HoMACANTHUs DELiCATULUS Eastman 

5126. Holotype. Kinderhook limestone. Le Grand, Iowa. Coll. 
A. H. Worthen, 1896. Eastman, 1903b, p. 218, pi. Ill, fig. 28, and 
pi. V, fig. 59; 1908, p. 152. 

Lamna sp. ind. 

5152. Two teeth. Pacific red clay. Albatross Station 3681. 
Eastman, 1903a, p. 186, pi. I, fig. 9, 10. 

5160. Two teeth. Pacific green mud. Albatross Station 4656. 
Eastman, 1906a, p. 80, pi. II, figs. 6, 7. 

Lamna (Odontaspis) bronnii Agassiz 

5197. Two cotypes. Agassiz, Poiss. Foss., 3, p. 297, pi. XXXVIIa, 
figs. 8, 9. 

See Odontaspis bronnii Agassiz. 

Lamna lata (Agassiz) 

5195. Holotype. Upper Cretaceous. Pietersberg, Maestricht. 
Coll. Bronn, 1859. Agassiz, Poiss. Foss., 3, p. 271 (1843), pi. XXXII, 
fig. 26 (1838), as Otodus latus. 

Lanarkia horrida Traquair 

5215. Plesiotype. Downtonian. Monks Burn, Lanarkshire, Scot- 
land. H. C. Stetson leg. and don., 1927. Stetson, 1931, p. 147, fig. 4. 

5216. Plesiotype. Downtonian. Seggholm, Ayrshire. D. Stitt 
leg. H. C. Stetson don., 1927. Stetson, 1931, p. 148. 

5239. Plesiotype. Downtonian. Seggholm. H. C. Stetson leg. 
and don., 1927. ' Stetson, 1931, p. 149. 

Lasanius problematicus Traquair 

1565. Plesiotype. Downtonian. Seggholm, Ayrshire. H. C. Stet- 
son and D. Tait'leg., 1926. Stetson. 1927, figs. 1, 2, pp. 250, 251. 



schevill: fossil types 79~ 

Lepidocottus multipinnatus (von Meyer) 

1587 (Orig. No. 1028; B.S.N.H. 3493). Holotype. Molasse. Un- 
terkirchberg, near Ulm, Germany. Coll. Eser, through Boston 
Society of Natural History, 1925. Von Meyer, 1848, p. 783, Gobius 
imdtipinnatvs; 1851a, p. 106, pi. XVII, fig. 1, as Cottus (?) multipin- 
natus. 

1586, 1592 (Orig. Nos. 1026, 1027; B.S.N.H. 3491, 3492). Two 
plesiotypes (cotypes of Gobius {f) conicus von Meyer). Molasse. 
Unterkirehberg. Coll. Eser, through Boston Society of Natural His- 
tory, 1925. Von Meyer, 1851, p. 80, as Gobius (f) conicus von Meyer; 
1851a, p. 109, as Cottus (?) conicus von Meyer; 1851a, p. 107, pi. XVI, 
figs. 7, 9, as Cottus brcvis Agassiz, crrore. No. 1592 cannot be found 
and has not been seen; its identification as this type is uncertain. 

5211 (Orig. No. 986; B.S.N.H. 3447). Plesiotype. Molasse. Un- 
terkirehberg. Coll. Eser, through Boston Society of Natural History, 
1925. Von Meyer, 1856, p. 27, pi. I, fig. 6, as Gobius f 

Woodward (1901, p. 584) refers this specimen to this species, but it 
agrees much better with L. papyraceus (Agassiz). 

Lepidocottus papyraceus (Agassiz) 

5262 (formerly M.C.Z. 2911). Holotype. Oligocene. Monte 
Viale, near Vicenza, Italy. Coll. Bronn, 1859. Agassiz, 1832, p. 137 
(nom. nud.); Poiss. Foss., 4, p. 187 (1839), pi. XXXII, fig. 1 (1839), 
as Cottus papyraceus Agassiz. 

Lepidosteus atrox Leidy 

5l(]/. Plesiotype. Eocene (Green River shales). Fossil, Wyo. / ^ 

D. C. Haddenham leg., 1899. Eastman, 1900, pp. 57-58; 1900b, pp. ' 

69-72, pi. I, fig. 2. 

Leptolepis bronnii Agassiz 

5069, 5175. Two genocotypes. Agassiz, 1832, p. 146. 
See Leptolepis coryphaenoides (Bronn). 

Leptolepis coryphaenoides (Bronn) 

5069, 5175 (Orig. Nos. 4a-b, 13; formerly M.C.Z. 2970-2971, 2565). 
Two genocotypes. Lias. The Baar, Baden, Germany. Coll. Bronn, 
1859, ex Coll. Althaus, 1826. Bronn, 1830, pp. 20, 22, 28, pi. I, fig. 1 
(synthetograph), as Cyprinus; Agassiz, 1832, p. 146, as L. bronnii 
Agassiz. 



80 bulletin: museum of comparative zoology 

Leuciscus gibbus von Meyer 

2004 (Orig. No. 1013; B.S.N.H. 3478). Holotype. Molasse. Un- 
terkirchberg, near Ulm, Germany. Coll. Eser, through Boston Society 
of Natural History, 1925. Von Meyer, 1851, p. 80; 1851a, p. 98, pi. 
XV, fig. 6. 

5210 (Orig. No. 1014; B.S.N.H. 3479). Plesiotype. Molasse. 
Unterkirchberg. Coll. Eser, through Boston Society of Natural His- 
tory, 1925. Von Meyer, 1856, p. 24, pi. I, fig. 2. 

Leuciscus papyraceus (Bronn) 

5258, 5260 (formerly M.C.Z. 2908 and 2914, 2922). Two cotypes. 
Oligocene. Geistinger Busch, Siebengebirge, Germany. Coll. Bronn, 
1859, ex Coll. Goldfuss, 1828. Bronn, 1828, pp. 377-381, pi. HI, fig. 
9 (synthetograph), as Cyprinus; Agassiz, 1832, p. 132; Poiss. Foss., 5, 
Pt. 2, p. 31 (1839), pi. LVI, figs. 1, 3-4 (1835). 

5259 (formerly M.C.Z. 2931). Cotype. Oligocene. Geistinger 
Busch. Coll. Bronn, 1859. Bronn, 1828, pp. 377-381, as Cyprinus; 
Agassiz, 1832, p. 132; Poiss. Foss., 5, Pt. 2, p. 31 (1839), pi. LVI, fig. 
2 (1835). 

The Bronn Collection also includes a small series of this species, of 
which at least part belongs to the original material, as is indicated by 
Bronn's old labels bearing the horizon and locality and the notation, 
" von Prof. Goldfuss in Tausch, 1828." x\nother label, however, shows 
that some of this material was not acquired by Bronn till 1830. There- 
fore it is possible that No. 5259, which was figured by Agassiz alone, 
may not be a cotype, but merely a plesiotype. 

\ 

Macropetalichthys agassizii (von Meyer) 

5174. Plesiotype. Woodward, 1891, p. 303; Eastman, 1907b, pp. 
101, 112; 1908, p. 175. 
See Macropetalichthys hoeninghausi (von Meyer). 

Macropetalichthys hoeninghausi (von Meyer) 

5174. Holotype (genoholotype of Physichthys hoeninghausi von 
Meyer). Devonian. Priim, Eifel, Germany. Kroffges leg., 1859. 
Coll. Schultze, 1871. Von Meyer, 1855, pp. 80, 83, pi. XV, figs. 1-5, as 
Physichthys; Woodward, 1890, p. 459; 1891, p. 303; Eastman, 1898, p. 
487; 1900c, p. 177; 1907b, pp. 101, 112; 1908, p. 175, as M. agassizii 
(von Meyer); Stetson, 1930. pp. 32-33, fig. 3. 

In the absence of a detailed description of the type of Macrope- 



schevill: fossil types 81 

talichtlujs agassizii (von Meyer), the name M. hoeninghausi is kept for 
our specimen (5174), although both may represent the same species. 
Cf. Stetson, 1930. p. 32. 

Macropetalichthys RAPHEiDOLABis Norwood and Owen 

1427. Plesiotype. Onondaga Hmestone. Lime Rock, Genesee 
County, N. Y. Coll. Ward, 1896. Dissecting cranium. Stetson, 
1930, p. 31, pi. VI, fig. 4, and pi. VII, figs. 1, 2. 

1428. Plesiotype. Onondaga limestone. Lime Rock, N. Y. Coll. 
Ward, 1896. Median posterior of cranium. Eastman, 1897b. p. 497, 
pi. XII, fig. 2, as M. sullimnti Newberry; 1907b, p. 107, fig. 21. 

2062 (Orig. No. M. No. 10). Plesiotype. Devonian. Sandusky, 
Ohio. Coll. A. A. Wright. Exch. Oberlin College Museum, 1896. 
Three detached sinistral plates. Eastman, 1897b, p. 497, pi. XII, fig. 
3, as M. suUhxinii Newberry. 

Macropetalichthys sullivanti Newberry 

1428, 2062. Two plesiotypes. Eastman, 1897b, p. 497, pi. XII, 
figs. 2, 3. 
See Macropetalichthys rapheidolabis Nor\\?ood and Owen. 

Megalichthys macropomus Cope 

5143. Plesiotype. Pennsylvanian. Lansing, Kans. Oscar Lamb 
leg., 1888. Coli O. H. St. John, 1900. Eastman, 1903b, p. 187. 

Myliobatis sp. 

5193. Eocene. Ashley River, S. C. Caudal spine. Eastman, 1901, 
pi. XIII, fig. 4. 

5194. Eocene (.Jackson group). Montgomery, La. Coll. F. V. 
Hopkins. Caudal spine. Eastman, 1901, pi. XIII, fig. 5. 

Myliobatis magister Leidy 

2063. Plesiotype. Eocene. Ashley River, S. C. Coll. Capt. 
BowTnan. Upper dentition. Eastman, 1901, p. 100, pi. XII, fig. 3, 
and pi. XIII, figs. la-b. 

Mylostoma newberryi Eastman 

1439 (Orig. No. 22). Cotype. Cleveland shale. Sheffield, Ohio. 
Coll. Terrell, 1885. Eastman, 1906, p. 23, pi. I, fig. 7, as M. variahile 
Newberry; 1907a, p. 224, fig. D (partim). 



82 bulletin: museum of comparative zoology 

Mylostoma terrelli Newberry 

1430 (Orig. No. 25). Holotype. Cleveland shale. Sheffield, Ohio. 
Coll. J. Terrell, 1885. Newberry, 1883, p. 147; 1889, p. 164, pi. XIV, 
figs. 1,2; Eastman, 1906, p. 23, pi. Ill, fig. 21. Hussakof (1908, p. 17) 
unites this species with M. variabile. 

Mylostoma variabile Newberry 

1429 (Orig. No. 26b). Genocotype. Cleveland shale. Sheffield, 
Ohio. Coll. J. Terrell, 1885. Newberry, 1883, p. 146; 1889, p. 165, 
pi. XV, figs. 1,2; Eastman, 1906, p. 23, pi. Ill, fig. 19. 

1431 (Orig. No. 26). Genocotype. Clevelajid shale. Sheffield, 
Ohio. Coll. J. Terrell, 1885. Newberry, 1883, p. 146; 1889, p. 165, 
pi. XV, fig. 3; Eastman, 1906, p. 23, pi. Ill, fig. 20. 

1435 (Orig. No. 27a). Genocotype. Cleveland shale. Sheffield, 
Ohio. Coll. J. Terrell, 1885. Newberry, 1883, p. 146; 1889, p. 165, 
pi. XV, figs. 4, 4a; Eastman, 1906, p. 23, pi. I, fig. 2. 

1436 (Orig. N^o. 27). Genocotype. Cleveland shale. Sheffield, 
Ohio. Coll.^J. Terrell, 1885. Newberry, 1883, p. 146; 1889; p. 165, 
pi. XV, figs. 5, 5a; Eastman, 1906, p. 23, pi. II, fig. 15. 

1437 (Orig. No. 27c). Plesiotype. Cleveland shale. Sheffield, 
Ohio. Coll. J. Terrell, 1885. Eastman, 1906, p. 23, pi. I, fig. 9; 
1907a, pi. (specimen at reader's right of group). 

1438 (Orig. No. 27b). Plesiotype. Cleveland shale. Sheffield, 
Ohio. Coll. J. Terrell, 1885. Eastman, 1906, p. 23, pi. II, fig. 11. 

1439(Orig. No. 22). Plesiotype. Eastman, 1906, p. 23, pi. I, fig. 7. 

See Mylostoma newberryi Eastman. 

1490. Plesiotype. Cleveland shale. Vicinity of Cleveland, Ohio. 
Coll. W. Kepler,\901. Dean, 1901, p. 101, pi. VIII; Eastman, 1907a. 
pp. 215 ff., figs. A, B, C. This specimen is the counterpart of No. 7526 
in the American Museum of Natural History. 

Odontaspis bronnii Agassiz 

5197. Two cotypes. Upper Cretaceous. Pietersberg, Maes- 
tricht. Coll. Bronn, 1859. Agassiz, Poiss. Foss., 3, p. 297 (1843), pi. 
XXXVIIa, figs. 8. 9 (1844), as Lamna (0.) hromvil. 

5196. Two plesio types (cotypes of Otodvs scrrafus Agassiz). Upper 
Cretaceous. Pietersberg, Maestricht. Coll. Bronn, 1859. Agassiz, 
Poiss. Foss., 3, p. 272 (1843), pi. XXXII, figs. 27, 28 (1838), as Otodvs 
serratus; Woodward, 1889, p. 360, 401. 



schevill: fossil types 83 

Onychodus sp. ind. 

5091. Devonian. Muhlenberg, near Gerolstein, Eifel, Germany. 
Coll. Sehultze, 1871. Woodward, 1891, p. 393; Eastman, 1907b, p. 169, 
pi. T, fig. 4; 1908, p. 240, pi. I, fig. 12 (| natural size, notf as noted). 

Onychodus hopkinsi Newberry 

5078 (Orig. No. 1199). Plesiotype. Chemung. Franklin, N. Y. 
Coll. Dyer, 1879. Eastman, 1899, p. 322, fig. 3 (error in text refers to 
fig. 2), as 0. sigmoides; 1907b, p. 169, pi. I, fig. 14. 

Onychodus sigmoides Newberry 

5078. Plesiotype. Eastman, 1899, p. 322, fig. 3 (error in text re- 
fers to fig. 2). 
See Onychodus hopkinsi Newberry. 

Orodus intermedius Eastman 

5120. Holotj'pe. Pennsylvanian. Vicinity of Weston, Mo. Coll. 
S. A. Miller, 1898. Eastman, 1903b, p. 183, pi. IV, figs. 35, 36. 

Otodus latus Agassiz 

5195. Holotype. Agassiz, Poiss. Foss., 3, p. 271 (1843). pi. XXXII, 
fig. 26 (1838). 

See Lamna lata (Agassiz). 

Otodus serratus Agassiz 

5196. Two cot\T)es. Agassiz, Poiss. Foss., 3, p. 272 (1843), pi. 
XXXII, figs. 27, 28 (1838). 

See Odontaspis bronnii Agassiz. 

Oxyrhina sp. 

5164. Pacific dark broA\Ti clay. Albatross Station 4701. East- 
man, 1906a, pp. 81, 82, pi. II, fig. 18. 

Oxy'rhina crass a Agassiz 

5145. Two plesiotypes. Tertiary phosphates. Coosaw, S. C. 
Coll. Capt. Bowman. Upper lateral teeth. Eastman, 1903a, p. 186, 
figs. 1-2, 3. 



84 bulletin: museum of comparative zoology 

5153. Three plesiotx-pes. Pacific red clay. Albatross Station 
3681. Anterior teeth. Eastman, 1903a, p. 185, pi. I, figs. 11-13. 

5154. Five plesiotypes. Pacific red clay. Albatross Station 
3681. Postero-lateral teeth. Eastman, 1903a, p. 185, pi. I, figs. 
14-18. 

5155. PlesiotN-pe. Pacific red clay. Albatross Station 3681. 
Lateral tooth. Eastman, 1903a, p. 185, pi. I, fig. 20. 

5156. Plesiot\'pe. Pacific red clay. Albatross Station 3683. 
Lateral tooth. Eastman, 1903a, p. 185, pi. I, fig. 19. 

5158. Three plesiotypes. Pacific radiolarian ooze. Albatross 
Station 4658. Postero-lateral teeth. Eastman, 1906a, pp. 80, 82, 
pi. II, figs. 1-3. 

5159. Two plesiot^'pes. Pacific green mud. Albatross Station 
4656. Postero-lateral teeth. Eastman, 1906a, pp. 80, 82, pi. II, 
figs. 4, 5. 

5161. Two plesiotypes. Pacific brown clay. Albatross Station 
4701. Postero-lateral teeth. Eastman, 1906a, pp. 81, 82, pi. II, 

figs. 8, 9. 

5162. Seven plesiotypes. Pacific brown clay. Albatross Sta- 
tion 4685. Eastman, 1906a, pp. 80, 82, pi. II, figs. 10-12, 14-17. 

Oxyrhina hastalts Agassiz 

5171, 5172. Two plesiotypes (cotypes of 0. .riphodon Agassiz). 
Miocene. Paris, France. Coll. Bronn, 1859. Agassiz, Poiss. Foss., 
3, p. 278 (1843), pi. XXXIII, figs. 11, 12, as 0. xiphodon. 

5173. Plesiotype (cotype of 0. 2>/fcrtft7?> Agassiz). Pliocene. Cas- 
tel-Arquato, Italy. Coll. Bronn, 1859. Agassiz, Poiss. Foss., 3, p. 
279 (1843), pi. XXXVII, figs. 14, 14a (1844) as 0. plicatilis. 

Oxyrhina plicatilis Agassiz 

5173. Cot>T)e. Agassiz, Poiss. Foss., 3, p. 279 (1843), pi. XXXVII, 
figs. 14, 14a (1844). 

See Oxyrhina hastalis Agassiz. 

Oxyrhina xiphodon Agassiz 

5171, 5172. Two cotypes. Agassiz, Poiss. Foss., 3, p. 278 (1843), 
pi. XXXIII, figs. 11, 12 (1838). 
See Oxyrhina hastalis Agassiz. 



schevill: fossil types 85 



Palaeomylus predator Eastman 

2050. Holotype. Devonian. Gerolstein, Eifel, Germany. Coll. 
Schultze, 1871. ' Eastman, 1898, p. 549, fig. 43, p. 483. 

Palaeoniscus sp. 

5086, 5087 (B.S.X.H. 7897, 7897a; formerly M.C.Z. 1959, 1958). 
Jackson, 1851, pp. 23, 24; Lambe, 1910, p. 17, pi. II, figs. 2, 2 bis, and 3. 

See Rhadinichthys alberti (Jackson). 

5088 (B.S.N.H. 7903; formerly M.C.Z. 1953). Jackson, 1851, p. 
24; Lambe, 1910, p. 17, pi. II, fig. 7. 

See Elonichthys brownii (Jackson). 

Palaeoniscus alberti Jackson 

5082 (B.S.N.H. 7899; formerly M.C.Z. 1960). Holotype. Jack- 
son, 1851, p. 22; 1852, p. 138; Lambe, 1910, pp. . 17, 21, pi. I, fig. 1. 

See Rh.ajoinichthys alberti (Jackson). 

Palaeoniscus brownii Jackson 

5083 (B.S.N.H. 7900; formerly M.C.Z. 1961). Holot>T)e. Jack- 
son, 1851, p. 22; 1852, p. 139; Lambe, 1910, pp. 17, 22, pi. I, fig. 2. 

See Elonichthys brownii (Jackson). 

Palaeoniscus cairnsii Jackson 

5084 (B.S.N.H. 7899a; formerly M.C.Z. 1956). Holotype. Jack- 
son, 1851, p. 22; 1852, p. 139; Lambe, 1910, pp. 17, 21, pi. I, fig. 3. 

See Rhadinichthys alberti (Jackson). 

Palaeophichthys parvulus Eastman 

5090. Genoholotype. Pennsylvanian. Mazon Creek, Grundy 
County, 111. Coll. S. S. Strong. Yale Peabody Museum don. East- 
man, 1908, p. 253, fig. 37 (figure is x3, not x2 as noted). 

Peripristis benniei (Etheridge, Jr.) 

270. Plesiotype. Carboniferous. Richmond, Yorkshire, Eng- 
land. Coll. L. Agassiz, 1859. Eastman, 1902a, p. 391, fig. 2, 
p. 390. 



86 bulletin: museum of comparative zoology 

Peripristis semicircuLuVRIS (Newberry and Worthen) 

5191. Plesiotype. Chester limestone. Montgomery Switch, Cald- 
well County, Ky. E. O. Ulrich leg., 1898. Lower tooth. Eastman, 
1902a, p. 389, figs, la-b, p. 390; 1903b, p. 179, fig. 7. 

5192. PlesiotN-pe. Chester limestone. Montgomery Switch, Cald- 
well County, Ky. E. O. Ulrich leg., 1898. Upper tooth. Eastman, 
1903b, p. 179, text fig. 8, pi. Ill, fig. 25. 

PETALORITiNCHUS PSITTACTNUS (McCoy) 

157. Plesiotype. Lower Carboniferous. Armagh, L-eland. Coll. 
L. G. de Koninck, 1861. Series of seven teeth. Eastman, 1903b, p. 
171, fig. 4. 

Phoebodus sp. 

5133. Eastman, 1899a, p. 491. 

See Phoebodus knightianus Eastman. 

Phoebodus dens-neptltni Eastman 

5132. Holotype. Keokuk limestone. Keokuk, Iowa. Coll. A. H. 
Worthen, 1896."^ Eastman, 1903b, p. 196, pi. IV, fig. 39. 

Phoebodus knightianus Eastman 

5133. Holotype and paratype. Permo-Carboniferous (Florence 
flint). Blue Springs, Nebr. \Y. C. Knight leg., 1898. Knight, 1899, 
pp. 366, 372, 374, as Diplodus sp.; Eastman^l903b, p. 169, pi. IV, figs. 

^^^^>f-^^^)/^ 40,40a. A 

Phoebodus politus Newberry 

5089. Plesiotype. Cleveland shale. Lorain County, Ohio. Coll. 
J. Terrell, 1885. Eastman, 1899a, p. 491, pi. VII, fig. 5; 1907b, p. 61, 
pi. I, fig. 12; 1908, p. 106, pi. I, fig. 9. 

Physichthys ? sp. 

1459, 1458. Two fragments. Von Meyer. 1855, pp. 81-82, pi. 
XV, figs. 6, 7-8. 

See Pterichthys rhenanus Beyrich. 

2059. Two fragments. Von IMeyer, 1855, p. 82, pi. XV, figs. 9, 
10-11. 

See Rhynchodus rostratus Eastman. 



schevill: fossil types 87 

Physichthys hoeninghausi von Meyer 

5174. Genoholotype. Von Meyer, 1855, pp. 80, 83, pi. XV, figs. 
1-5. 
See Macropetalichthys hoeninghausi (von Meyer). 

Physonemus arcuatus McCoy 

1550. Plesiotype. Mississippian (St. Louis limestone). Alton, 
111. Coll. A. H. Worthen, 189G. Eastman, 1903b, p. 208, fig. 12. 

Physonemus hamus-piscatorius Eastman 

5125. Two cotypes. Kinderhook limestone. Burlington, Iowa. 
Coll. C. Wachsmuth, 1872. Eastman, 1903b, p. 207, pi. V, figs. 45, 46. 

Physonemus pandatus Eastman 

5140 (Orig. No. 54). Holot^-pe. Kinderhook limestone. Bogus 
Hollow, Burlington, Iowa. Coll. C. Wachsmuth, 1872. Eastman, 
1903b, p. 207, pi. V, fig. 44. 

Prionodon ? denticulatus Agassiz 

5077. Agassiz, 1856, p. 274. 

See Galeocerdo denticulatus Agassiz. 

Protitanichthys fossatus Eastman 

5220. Genoholotv'pe. Delaware limestone. Delaware, Ohio. 
Coll. H. Herzer, 1898. Eastman, 1907b, p. 144, fig. 30, p. 146, pi. X, 
fig. 2; Hussakof, 1908a, p. 311, as Coccosteus? 

Pterichthys rhenanus Be;yTich 

1458. Plesiotv-pe. Devonian. Priim, Eifel, Germany. KrofTges 
leg., 1858. Coll. Schultze, 1871. Anterior median dorsal. Von 
Meyer, 1855, p. 82, pi. XV, figs. 7-8, as Physichthys?; Woodward, 
1890, p. 459; 1891, p. 222; Eastman, 1898, p. 487. 

1459. Plesiotype. Devonian. Priim. Coll. Schultze, 1871. Left 
posterior ventro-lateral. Von Meyer, 1855, p. 81, pi. XV, fig. 6, as 
Physichthys?; Woodward, 1890, p. 459; Eastman, 1898, p. 487. 

Ptyctodus calceolus Newberry and Worthen 

2054. PlesiotA'pe. L^pper Devonian (state quarry beds). North 
Liberty, Iowa. Coll. Calvin, 1897. Eastman, 1898, p. 477, fig. 15. 



88 bulletin: museum of comparative zoology 

2055. Thirteen plesiotypes. State quarry beds. North Liberty, 
Iowa. Coll. Calvin, 1897. Eastman, 1898, pp. 477-478, figs. 1-10,14, 
16-17. 

2056. Three plesiotypes. State quarry beds. North Liberty, 
Iowa. Coll. Calvin, 1897. Eastman, 1898,' p. 478, figs. 11-13, p. 477. 

Ptyctodus compressus Eastman 

2052. Twocot^'pes. L'pper Devonian (state quarry beds). North 
Liberty, Iowa. C. R. Eastman leg., 1897. Eastman, 1898, p. 479, 
figs. 21, 23, p. 477. 

2053. Eight cotypes. State quarry Vjeds. North Liberty, Iowa. 
C. R. Eastman leg.," 1897. Eastman, 1898, p. 479, figs. 18-20, 22, 24- 
27, p. 477. 

Ptyctodus ferox Eastman 

2051. Holotype. Mid-Devonian (Hamilton). Milwaukee, Wis. 
E. E. Teller leg.*^ and don., 1898. Eastman, 1898, p. 480, fig.35; 1908, 
p. 136, fig. 22. 

Ptyctodus molaris Eastman 

2048. Holotype. Devonian. Priim, Eifel, Germany. Kroffges 
leg., 1859. Coll. Schultze, 1871. Eastman, 1897a, p. 115, fig. lOB; 
1898, p. 475, fig. 28, p 477. 

2049. Two paratypes. Devonian. Eifel. Coll. Schultze, 1871. 
Eastman, 1897a, p. 115; 1898, p. 475, figs. 29, 30, p. 477. 

Ptyctodus panderi Eastman 

204G. Holotype. Devonian. Berndorf, Eifel, Germany. KrolTges 
leg., 1859. Coll Schultze, 1S71. Eastman, 1898, p. 484, fig. 32, p. 477. 

2045. Three paratypes. Devonian. Gerolstein. Eifel. Kroffges 
leg., 1859. Coll. Schultze, 1871. Eastman, 1898, p. 484, figs. 31, 33, 
34, p. 477. 

Ptyctodus punctatus Eastman 

2058. Holotype. Devonian (Onondaga). Le Roy, N. Y. Coll. 
Ward, 1901. Eastman, 1907h, p. 70, fig. 15a (the figure is not natural 
size, as noted, but is nearly twice that); 1908, p. 133, pi. Ill, fig. 6. 

Pygaeus agassizii Eastman 

5073. Holotype. Eocene. Monte Bolca, near Verona, Italy. 
Coll. Canossa, through Krantz, 1903. Eastman, 1904, p. 31, pi. II. 



schevill: fossil types 89 

Pygaeus nobilis Agassiz 

5264. Agassiz, 1835, pp. 12, 20, 29, 39; 1835a, pp. 295, 302, 309, 
313; Poiss. Foss., 4, p. 16* (1838), p. 253 (1842), errore. 
See Ac.oiTHURUS ovalis Agassiz. 

Rhadinichthys ? sp. 

5114 (formerly M.C.Z. 1620). Chemung. Warren, Pa. Coll. F. A. 
Randall, 1899. Eastman, 1907b, p. 172, pi. IV, figs. 10, 11, and pi. IX, 
fig. 4; 1908, p. 260. 

Rhadinichthys alberti (Jackson) 

5082 (B.S.N.H. 7899; formerly M.C.Z. 1960). Holotype. Mis- 
sissippian. Albert Mines, New Brunswick. Exch. Boston Society of 
Natural History, 1897. Jackson, 1851, p. 22; 1852, p. 138, as Palaeon- 
iscus; Lambe, 1909, p. 168; 1910, p. 21, pi. I, fig. 1. 

5084 (B.S.N.H. 7899a; formerly M.C.Z. 1956). Plesiotype (holo- 
type of Palaeoniscus cairtisii Jackson) . Mississippian. Albert Mines. 
Exch. Boston Society of Natural History, 1897. Jackson, 1851, p. 
22; 1852, p. 139, as Palaeoniscus cairnsii Jackson; Eastman, 1908, p. 
262, as Rhadinichthys cairnsii (Jackson); Lambe, 1909, p. 168, fig. 2, p. 
173; 1910, p^21, pi. I, fig. 3, and pi. Ill, fig. 4. 

5086 (B.s(n.H. 7897; formerly M.C.Z. 1959). Plesiotype. Mis- 
sissippian. Albert Mines. Exch. Boston Society of Natural History, 
1897. Jackson, 1851, p. 23, as Palaeoniscus sp.; Lambe, 1910, pi. II, 
fig. 2, 2bis. 

5087 (B.S.N.H. 7897a; formerly M.C.Z. 1958). Plesiot^-pe. Mis- 
sissippian. Albert Mines. Exch. Boston Society of Natural History, 
1897. Jackson, 1851, p. 24, as Palaeoniscus sp.; Lambe, 1909, p. 168, 
fig. 1 (?), p. 173 (cf. Lambe, 1910, pi. Ill, fig. 5); 1910, pi. II, fig. 3, 
and pi. Ill, figs. 5, 6. 

Rhadinichthys cairnsii (Jackson) 

5084 (B.S.N.H. 7899a; formerly M.C.Z. 1956). Holotype. East- 
man, 1908, p. 262. 
See Rhadinichthys alberti (Jackson). 

Rhadinichthys deani Eastman 

5222. Cot^^e. Waverly series. Vicinity of Junction City, Boyle 
County, Ky. Moritz Fischer leg., 1907. Eastman, 1908, pp. 264, 
267, figs. 40A-A1, and pi. XIII, figs. 8, 9; Parker, 1908, p. 272. 



19- 



90 bulletin: museum of comparative zoology 

5223, 5225-5236. Thirteen cotypes. Waverly series. Boyle County, 
Ky. Moritz Fischer leg., 1907. Eastman, 1908, pp. 264, 267, 268, 
figs. 40, 41, and pi. XIII. 

5224. Cotype. Waverly series. Boyle County, Ky. Birdie 
Linney leg., 1907. Eastman, 1908, fig. 40c, p. 267. 

5257. Cotype. Waverly series. Boyle County, Ky. Moritz 
Fischer leg., 1907. Eastman, 1908, pp. 264, 266. 

Rhombus kirchberganus von Mever 

2000. Cotype. Von Meyer, 1848, p. 782. 
See SoLEA KiRCHBERGANA von Meyer. 

Rhynchodus sp. ind. 

2059-A. Woodward, 1891, p. 39. 

See Rhynchodus rostratus Eastman. 

Rhynchodus major Eastman 

2042. Holotype. Devonian. Berndorf, near Kerpen, Eifel, Ger- 
many. Coll. Schultze, 1871. Eastman, 1898, p. 487, fig. 42, p. 483. 

Rhynchodus rostratus Eastman 

2043. Five cotjT)es. Devonian. Eifel, Germanv. Coll. Schultze, 
1871. Eastman, 1898, p. 487, figs. 41, 44-47, p. 483. 

2059. Two cotypes. Devonian. Priim, Eifel. Kroffges leg., 
1859. Coll. Schultze, 1871. Von Meyer. 1855, p. 82, pi. XV, figs. 9, 
l/ooJwar<i,. 10-11' as Physichthys?; 1891, p. 39; Eastman, 1898, p. 487. 

Semionotus nilssoni Agassiz 

5067 (formerly M.C.Z. 2685). Holotype. Upper Triassic. . Scania, 
Sweden. Nilsson, 1824, p. 103, pi. II, fig. 1, as an indeterminate 
acanthopterygian; Agassiz, Poiss. Foss., 2, Pt. 1, p. 229 (1837), pi. 
XXVIIa, figs. 1-5 (1844); Eastman, 1905, p. 74, figs. 10, 11; 1911, p. 
58, pi. VI. 

Smerdis elongatus von Meyer 

1593 (Orig. No. 1000; B.S.N.H. 3465). Holotj-pe. Molasse. Un- 
terkirchberg, near Ulm, Germany. Coll. Eser, through Boston 
Society of Natural History. 1925. Von Meyer, 1851, p. 80; 1851a, p. 
110, pi. XVI, fig. 6. 



schevill: fossil types 91 



Smerdis formosus von Mever 

2003 (Orig. No. 992; B.S.X.H. 3453). Four cotypes. Molasse. 
Unterkirchberg, near Ulm, Germany. Coll. Eser^ through Boston 
Society of Natural History, 1925. Von Meyer, 1848, p. 783; 1851a, 
p. 110, pi. XVI, fig. 5; Cushman, 1907, p. 271, as holotype. 

Smerdis minutus (Blain\alle) 

1594 (Orig. No. 994; B.S.N.H. 3455). Plesiotype. Molasse. Un- 
terkirchberg, near Ulm, Germany. Coll. Eser, through Boston So- 
ciety of Natural History, 1925. Von Meyer, 1848, p. 783; 1851a, p. 
109,^ pi. XVI, figs. 1, 2. 

1595 (Orig. No. 993; B.S.N.H. 3454). Two plesiotypes. Molasse. 
Unterkirchberg. Coll. Eser, through Boston Society of Natural His- 
tory, 1925. Von Meyer, 1848, p. 783; 1851a, p. 109, pi. XVI, fig. 3. 

Von Meyer credits the species to Agassiz. 

Solea antiqua von Meyer 

2005 (B.S.N.H. 3486). Holotype. Von Meyer, 1851, p. 80; 1851a, 
p. 103, pi. XVII, figs. 4,5,6, 7; Cushman, 1907, p. 272, as cotype, errore. 

1596 (B.S.N.H. 3484). Cushman, 1907, p. 272, as cotype, errore. 

1598 (B.S.N.H. 3487). Plesiotype. Von Meyer, 1856, p. 26, pi. I, 
fig. 4. 

5212. Plesiotype. Von Meyer, 1856, p. 26, pi. I, fig. 5. 
See Solea kirchbergana von Meyer. 

Solea kirchbergana von Meyer 

1599 (Orig. No. 1020; B.S.N.H. 3485). Cushman, 1907, p. 272, as 
cot^^pe, errore. 

2000 (Orig. No. 1023; B.S.N.H. 3488). Cotype. Molasse. Un- 
terkirchberg, near Ulm, Germany. Coll. Eser, through Boston Society 
of Natural History, 1925. Von Meyer, 1848, p. 782, as RJwmbvs kirch- 
berganus; 1851, p. 80; 1851a, p. 102, pi. XVII, fig. 3. 

1596 (Orig. No. 1019; B.S.N.H. 3484). Plesiotype. Molasse. 
Unterkirchberg. Coll. Eser, through Boston Society of Natural His- 
tory, 1925. Von Meyer, 1856, p. 25, pi. I, fig. 3; Cushman, 1907, p. 
272, as cotype of S. aniiqva von ]\Ieyer, errore. 

1598 (Orig. No. 1022; B.S.N.k. 3487). Plesiotype. Molasse. 
Unterkirchberg. Coll. Eser, through Boston Society of Natural His- 



92 bulletin: museum of comparative zoology 

tory, 1925. Von Meyer, 1856, p. 26, pi. I, fig. 4, as S. anfiqtta von 
Meyer. 

2005 (Orig. No. 1021; B.S.X.H. 3486). Plesiotype (holotype of S. 
antiqua von yieyer). Molasse. Unterkirchberg. Coll. Eser, through 
Boston Society of Natural History, 1925. Von Meyer, 1851, p. 80; 
1851a, p. 103, pi. XVII, figs. 4, 5, 6, 7, as S. antiqua von Meyer; Cush- 
man, 1907, p. 272, as cotype of S. antiqua, crrore. 

5212(Orig. No. 1025; B^S.X.H. 3490). Plesiotype. Molasse. Un- 
terkirchberg. Coll. Eser, through Boston Society of Xatural History, 
1925. Von Meyer, 1856, p. 26, pi. I, fig. 5, as S. antiqua von Meyer. 

Stethacanthus depressus (St. John and Worthen) 

• dt/^^toi 5136. Plesiotype. Waverly sandstone. Marshal LCounty, Mich. 

/^ Coll. W. D. Gunning. Eastman, 1903b, p. 216, fig. 15 (poor figure) ; 

Hussakof, 1913, p. 249, footnote, and Hussakof and Bryant 1919, p. 



170, footnote, as S. humilis Hussakof, error 



Stethacanthus humilis Hussakof 

5136. Plesiotype. Hussakof, 1913. p. 249, footnote, errore; Hus- 
sakof and Bryant, 1919, p. 170, footnote, errore; 

See Stethacanthus depressus (St. John and ^Vorthen). 

Symphodus szajnochae (de Zigno) 

5072. (Orig. Nos. 17, 18; formerly M.C.Z. 3078). Plesiotv-pe. 
Eocene. Monte Bolca, near Verona. Italy. Coll. Krantz, 1903. 
Eastman. 1904. p. 29, pi. I. fig. 5. 

Synthetodus calvini Eastman 

5098. Holotype. Upper Devonian (state quarry beds). North 
Liberty, Johnson County, Iowa. Coll. Calvin, 1896. Eastman, 
1907b, pi. IV, fig. 15, p. 203, as Dipnoan dental plate, gen. et sp. nov. ; 
1908, p. 233, pi. II, fig. 19. 

5099. Fifteen paratypes. State quarry beds. Johnson and IMus- 
catine Counties, Iowa. Coll. Calvin, 1897. Eastman, 1908, p. 233, 
pi. XII, figs. 1-15. 

Taeniodus contortus St. John and Worthen 

5137. GenoholotN-pe. St. John and Worthen, 1883, p. 76. 
See Deltodus contortus (St. John and Worthen) . 



schevill: fossil types 93 

Thelodus macintoshi Stetson 

2035. Holotype. Upper Silurian. Cunninghams Mill Brook. 
Nerepis, New Brunswick. H. C. Stetson leg. and don., 1927. Stetson, 
192Sb, p. 223, fig. l,p. 222. 

2036, 2037. Two paratypes. Upper Silurian. Cunninghams Mill 
Brook, Nerepis, New Brunswick. H. C. Stetson leg. and don., 1927. 
Stetson, 1928b, p. 223, figs. 2-3, 4-6, p. 222. 

Thelodus planus Traquair 

5217. Plesiotype. Ludlow. Logan Water, Lanarkshire, Scot- 
land. H. C. Stetson leg. and don. Stetson, 1931, p. 148. 

5250. Plesiotype. Ludlow. Logan Water. D. Tait leg. 1929. 
Stetson, 1931, p. 143, fig. 2A, p. 144. 

Thelodus scoticus Traquair 

5255. Plesiot^'pe. Ludlow. Logan Water, Lanarkshire, Scot- 
land. H. C. Stetson leg. and don., 1928. Stetson, 1931, p. 142, figs. 
lA, IB. 

Thelodus taiti Stetson 

5248. Paratype. Downtonian. ^Vlonks Burn, Lanarkshire, Scot- 
land. D. Tait leg., 1929. Stetson, 1931, p. 143, figs. IC, ID, p. 142. 
and fig. 2C, p. 144. 

5249. Paratope Downtonian. Monks Burn. D. Tait, P. E. 
Raymond, and H. C. Stetson leg., 1928. Stetson, 1931, p. 143, fig. IE. 
p. 142, and fig. 2B, p. 144. 

TiTANICHTHYS sp. 

1301. Antero-ventro-median. Eastman, 1896, p. 47. 

See DiNiCHTHYS terrelli Newberry. 

1300. Postero-ventro-median. Eastman, 1896, p. 47. 

See DiNiCHTHYS sp. 

1475. Postero-ventro-median. Eastman, 1907b, p. 144. 1908, p. 205. 

See under Diniehthyid. 

TiTANICHTHYS AGASSizii Newberry 

1297 (Orig. No. 1). Genoholotype. Cleveland shale. Sheffield, 
Ohio. Coll. J. Terrell, 1885. Newberry, 1885, p. 27; 1888, p. 13' 
genus only; 1899, p. 133, pi. I, pi. IV, fig. 4; Eastman, 1898, p. 761, 
fig. 4, p. 763. 



94 bulletin: museum of comparative zoology 



Tristychius fimbriatus Stock 

5201 (formerly M.C.Z. 4259). Holotype. Stock, 1883, p. 177, pi. 
VII, figs. 1, la. 

See Harpacanthus fimbriatus (Stock). 

Tristychius arcuatus Agassiz 

5202 (formerly M.C.Z. 4255). Plesiotype. Lower Carboniferous 
(Calciferous). Carolina Park, Edinburgh, Scotland. Coll. T. Stock, 
1883. Associated spine and axial skeleton. Stock, 1883, p. 180, pi. 
VII, fig. 8. 

5203 (formerly M.C.Z. 4263). Plesiotype. Calciferous. Trinity, 
near Edinburgh. Coll. T. Stock, 1883. Four associated teeth. 
Stock, 1883, p. 183, pi. VII, fig. 12. 

5204 (formerly M.C.Z. 4264). Plesiotype. Calciferous. Trinity. 
Coll. T. Stock, 1883. Tooth. Stock, 1883. p. 187, pi. VII, fig. 18. 

5205 (formerly M.C.Z. 4265). Plesiotype. Calciferous. Hailes 
quarry, near Edinburgh. Coll. T. Stock, 1883. Tooth. Stock, 
1883, V- 182. pi. VII, fig. 9. 

5207. Plesiotype. Calciferous. Bathgate. Coll. T. Stock, 1883. 
Tooth. Stock, 1883, p. 183, pi. VII, fig. 11. 

5208. Plesiotype. Calciferous. Hailes quarry. Coll. T. Stock, 
1883. Tooth. Stock, 1883, p. 182, pi. VII, fig. 10. 

5209 (formerly M.C.Z. 4261). Plesiotype. Calciferous. Hailes 
quarry. Coll. T. Stock, 1883. Spine. Stock, 1883, p. 184, pi. VII, 
figs. 13, 13a. 

Vertebral centrum 

5100. Upper Devonian (state quarry beds). Solon, Iowa. J. R. 
Hoats leg. and don., 1907. Eastman, 1908, p. 147, pi. XII, fig. 16. 

Wardichthys cyclosoma Traquair 

5246 (formerly M.C.Z. 3378). Plesiotype. Lower Carboniferous 
(Calciferous). Wardie, Edinburgh, Scotland. Coll. T. Stock, 1883. 
Stock, 1881, p. 490. 

XiPHOTRYGON ACUTIDENS CopC 

2038. Genoholotype. Eocene (Green River shales). Twin Creek, 
Bear River region, southwestern Wyoming. L. A. Lee leg., A. Agassiz 
don., 1910. Cope, 1879, p. 333; 1884. pp. 49-51, pi. I, figs. 1-5. 



schevill: fossil types 95 



REPTILES AND MIPHIBIANS 

Alligator prenasalis (Loomis) 

1014 (formerly M.C.Z 21699). Plesiotype. Oligocene (White 
River Titanotherium beds). Little Corral Draw, near Scenic, S. Dak. 
H. and P]. Schlaikjer leg., 1925. T. Barbour don. Incomplete skele- 
ton. Barbour, 1926. pp. 109-111. Mook, 1932, pp. 19-41; pi. 1-3. 

1015. Plesiotype. Oligocene (White River Titanotherium beds). 
Little Corral Draw, near Scenic, S. Dak. H. and E. Schlaikjer leg., 
1925. T. Barbour don. Skull and jaws with a few other bones. 
Barbour, 1926, pp. 109-111. Mook, 1932, pp. 19-41; - pi. 1- ^ . 

Archaeophis bolcensis IMassalongo 

1001, 1002 (Orig. Xos. 11, 12). Two cotypes. Eocene. Monte 
Bolca, near Verona, Italy. Coll. Canossa, through Krantz, 1903. 
Massalongo, 1859, pp. 15-16, pi. Ill, IV; Janensch, 1906, pp. 17-18. 

1003 (Orig. No. 13). Cotv-pe. Eocene. Monte Bolca. Coll. 
Canossa, through Krantz, 1903. Massalongo, 1859, pp. 15-16; 
Janensch, 1906, pp. 17-18. 

Baphetes minor Dawson 

1053. Holotype. Carboniferous. Vicinity of Joggins, Nova Scotia. 
H. A. Morrell leg. Mrs. E. M. Hooper don. Exterior of right man- 
dible. Dawson, 1870, p. 166; 1870a, p. 87; 1870b, p. 99. 

Batrachichnus plainvillensis Woodworth 

1052. Genoholotv-pe. Pennsylvanian. Plainville, Mass. J. B. 
Woodworth leg., 1899, et don., 1917. Woodworth, 1900. pp. 452-453, 
text fig. 2, pi. XL, fig. 1. 

The indeterminate scratches mentioned by Woodworth (1900, p. 454, 
pi. XL, fig. 2) are also in the collection. 

Belodon planirostris von Meyer 

1018 (Orig. No. 4361; B.S.N.H. 7512). Cotype. Triassic (Stuben- 
sandstein). Aixheim, near Spaichingen, Wiirttemberg, Germany. 
Coll. Eser, through Boston Society of Natural History. Fragmentary 
skull. Von Meyer, 1863, pp. 241-244, pi. XLI, figs. 1-3, and pi. XLII, 

fig. 7. 



J? 



96 bulletin: museum of comparative zoology 

1019 (Orig. No. 4369; B.S.N.H. 7520). Three cotypes. Stuben- 
sandstein. Aixheim. Coll. Eser, through Boston Society of Natural 
History. Snout fragments. Von Meyer, 1863, p. 243, pi. XLI, figs. 
4-9, and pi. XLII, fig. 7 (only 1019-A and 1019-B figured). 

1020 (Orig. No. 4365; B.S.N.H. 7516). Cotype. Stubensandstein. 
Aixheim. Coll. Eser, through Boston Society of Natural History. 
Dermal scute. Von Meyer, 1863, p. 244. pi. XLI, figs. 10-11. 

1021 (Orig. No. 4364; B.S.N.H. 7515). Cotype. Stubensandstein. 
Aixheim. Coll. Eser. through Boston Society of Natural History. 
Left femur. Von Meyer, 1863, p. 241. 

1022 (Orig. No. 4367; B.S.N.H. 7518). Two cotypes. Stubensand- 
stein. Aixheim. Coll. Eser, through Boston Society of Natural His- 
tory. Skull fragments. Von Meyer, 1863, p. 244. 

1023 (Orig. No. 4362; B.S.N.H.' 7513). Cotype. Stubensandstein. 
Aixheim. Coll. Eser, through Boston Society of Natural History. 
Dermal scute. Von Meyer, 1863, p. 244. 

This is the t^-pe species of Mystriosuchus E. Fraas, 1896. 

Carolinochelys wilsoni Hay 

1005 (Orig. No. 13640; formerly M.C.Z. 20300). Genoholotype. 
Eocene (Ligleside marl). Vicinity of Charleston, S. C. Coll. R. 
Wilson. T. Barbour don., 1924. Skull, humerus. Hay, 1923, pp. 
119-120; 1923a, pp. 29-31, pis. U, HL 

Chonespondylus grandis Leidy 

1049. Genoholotype. Leidy, 1S6S, p. 178. 
See Cymbospondylus petrinus Leidy. 

Crocodylus robustus Vaillant and Grandidier 

1006. Plesiotype. Pleistocene. Antsirabe, Madagascar. F. R. 
Wulsin leg. and don., 1917. Skull. Barbour, 1918, p. 488, pi. I, 
figs. 1-3. 

Cyclura sp. 

1016. Two jaws, one maxilla. Barbour, 1917, p. 98. 
See Cyclura portoricensis Barbour. 

Cyclura portoricensis Barbour 

1008 (formerly M.C.Z. 12460). Holotype. Pleistocene. Ciales 
Cave, Porto Rico. G. ]\L Allen and J. L. Peters leg., T. Barbour don., 



schevill: fossil types 



97 



1917. Left humerus (shaft missing). Barbour, 1919, p. 146, pi. I, 
figs. F, G; Barbour and Loveridge, 1929, p. 249. 

1009-1013 (formerly M.C.Z. 16870-16874). Five paratypes. Pleis- 
tocene. Ciales Cave, Porto Rico. G. M. Allen and J. L. Peters leg., 
T. Barbour don.. 1917. Left femur, tibia, ulna, right ulna, first sacral 
vertebra. Barbour, 1919, p. 146, pi. I, figs. A-E; Barbour and Lover- 
idge, 1929, p. 249, referring to 1010 (formerly 16871) only. 

1016. Three paratypes. Pleistocene. Ciales Cave, Porto Rico. 
G. M. Allen and J. L. Peters leg., T. Barbour don., 1917. Two jaws, 
one maxilla. Barbour, 1917, p. 98, as Cyclura sp.; 1919, pp. 145, 146. 

1017. Nine paratypes. Pleistocene. Ciales Cave, Porto Rico. 
G. M. Allen and J. L. Peters leg , T. Barbour don., 1917. Fragmental 
bones. Barbour, 1919, p. 146. 

Cymbospondylus GRANDis (Lcidy) 

1049. Holotype. Merriam, 1902, pp. 106-107. pi. XVI, fig. 3. 
See Cymbospondylus petrinus Leidy. 

Cymbospondylus petrinus Leidy 

1044(Orig.Nos. 68-72). Holotype. Middle Triassic. Humboldt. 
Nev. Coll. J. D. Whitney, 1896. Three fragmentary vertebrae and 
two plugs. Leidy, 1868, p. 178; Merriam, 1902, p. 106, pi. XVL figs. 
4, 5. The figure is of No. 1044-E (Orig. No. 72). 

1049 (Orig. No. 23). Plesiotype (genoholotype of C hones pondylus 
grandis Leidy). Middle Triassic. Star Canyon, Humboldt County, 
Nev. Coll. J. D. Whitney, 1896. Fragmentary centrum. Leidy, 
1868, p. 178, as C hones pondylus grandis Leidy; Merriam, 1902, pp. 
106-107, pi. XVI, fig. 3, as Cymbospondylus grandis (Leidy); 1908, p. 
104. 

Cymbospondylus piscosus Leidy 

1045 (Orig. No. 73). Genoholotype. Middle Triassic. New Pass, 
Toiyabe Range, Nev. Coll. J. D. Whitney, 1896. Five fragmentary 
associated vertebrae. Leidy. 1868, p. 177; Merriam, 1902. pp. 104- 
105, pi. XVI, figs. 1, 2; 1908, pp. 104, 123-124, figs. 136, 137. 

1047 (Orig. No. 73a). Genoparatype. Middle Triassic. Star 
Canyon, Humboldt County, Nev. Coll J. D. Whitney, 1896. Nine 
fragmentary associated centra. Leidy, 1868, p. 177. 

1048 (Orig. No. 73b). Genoparatype. Middle Triassic. Reese 
River, Toiyabe Range, Nev. Coll. J. D. Whitney, 1896. Vertebral 
centrum. Leidy, 1868, p. 178. 



98 bulletin: museum of comparative zoology 

Perhaps these specimens should be called cotypes ; in view of Leidy's 
wording, and since Merriam took No. 1045 as the lectotype, they are 
listed as above. 

DrOMOPUS ? WOODWORTHI Lull 

1050. Holotype. Pennsylvanian. Attleboro, Mass. F. Garnjost 
leg., 1916. J. B. Woodwortii don. Lull, 1920, pp. 234-236, fig. 1. 

HoMOEosAURUS MAXiMiLiANi von Meyer 

1004. Plesiotype. Upper Jurassic. Solenhofen, Bavaria, Ger- 
many. Coll. Haeberlein, 1882. Shows squam.ation. Barbour and 
Stetson, 1929, pp. 99-104, pi. I, figs. 1-2. 

LlODON PRORIGER Cope 

Holotype. Cope, 1870, pp. 202-205, pi. XII, figs. 22-24; 1875, pp. 
161-166, pi. XXX, figs. 10-14. 
See Tylosaurus proriger (Cope). 

Machaeroprosopus lithodendrorum Camp 

1029 (Orig. No. 7034/26719). Paratype. Triassic (Lower Chinle). 
Carrizo Wash, near Adamana, Ariz. Exch. University of Cali- 
fornia, 1931. Skull lacking rostrum. Camp, 1930, pp. 19, 26, 27, 29, 
30, 46-47, figs. 2a. 3d, 4c, 6, 13. 

Macrosaurus proriger Cope 

Holotype. Cope, 1869, p. 123; 1869a, pp. 121-122. 
See Tylosaurus proriger (Cope). 

Mystriosuchus planirostris (von Meyer) 

1018-1023. Nine genocotypes. 

See Belodon planirostris von Meyer. 

Plesiosaurus longirostris Blake 

1033. Holotj^e. Upper Lias. Whitby, Yorkshire. Brown- 
Marshall leg. Exch. Tufts College, 1931. Skeleton. Blake, 1876, 
p. 250-252, pi. I, fig. 6, and pi. Ill, fig. 2. 



schevill: fossil types 99 

Rana jageri von Meyer 

1007 (Orig. No. 479; B.S.N.H. 2765). Holotype. Miocene. Has- 
lach, near Ulm, Germany. Coll. Eser, through Boston Society of 
Natural History, 1925. Jager, 1850, p. 822, pi. LXXII, fig. 63 (re- 
composed figure), as a mole; von Meyer, 1851, p. 78; 1860, p. 144, pi. 
XXII, fig. 5. 

 Von Meyer hints (1860, p. 144) that Jager had only the counterpart 
of our specimens; Jager 's figure does not bear him out. 

Tylosaurus proriger (Cope) 

Holotype. Cretaceous (Niobrara). Near Monument, Kans. Colonel 
Connyngham and Mr. Minor leg. and don. Cope, 1869, p. 123, and 
1869a, pp. 121-122, as Macrosaurus; 1870, pp. 202-205, pi. XII, figs. 
22-24, and 1875, pp. 161-166, pi. XXX, figs. 10-14, as Liodon. 

This specimen was apparently never returned by Cope; nothing has 
been learned of its present whereabouts. 



BIRDS 

AcciPiTRiD, gen. et sp. ind. 

2219. Lower Miocene. Stenomylus quarry. Agate, Nebr. E. M. 
Schlaikjer leg., 1929. T. Barbour don. Broken right ulna, pedal 
phalanx, broken vertebra. Wetmore, 1930, p. 152. 

Branta canadensis hutchinsi (Richardson) 

2226. Plesiotype. Pleistocene. Melbourne, Fla. C. P. Single- 
ton leg., 1928. T. Barbour don. Right humerus. Wetmore, 1931, 
p. 19. 

Carinate contour feather 

2218 (formerly M.C.Z.%99). Eocene. Monte Bolca, near Verona, TVi 
Italy. Coll. Canossa, through Krantz, 1903. Eastman, 1904a, pp. 
669-672, fig., p. 671. 

Casmerodius albus (Linnaeus) 

2224. Plesiotype. Pleistocene. Melbourne, Fla. C. P. Single- 
ton leg., 1928. T. Barbour don. Right ulna. Wetmore, 1931, p. 
15. 



M 



100 bulletin: museum of comparative zoology 

Cathartes aura septentrionalis Wied 

2229. Plesiotype. Pleistocene. Melbourne, Fla. C. P. Single- 
ton leg., 1928. T. Barbour don. Distal end of left ulna. Wetmore, 
1931, p. 24. 

Gallinuloides wyomingensis Eastman 

2221 (formerly M.C.Z. 159S). Genoholotype. Eocene (Greei/ 
River shales). Fossil, Wyoming. D. C. Haddenham leg., 1899. 
Eastman, 1900, pp. 55-57, pi. IV; Lucas, 1900, pp. 79-84, text fig. and 
pi. I; Shufeldt, 1915, pp. 619-634, figs. 1, 2. 

Shufeldt's generic name "Palaeobonasa" (1915, p. 633) is invalid; 
unfortunately, it has been accepted by Lambrecht (1916, pp. 234, 296, 
495, and 1921, p. 80). 

Grus AMERICANA (Linnacus) 

2228. Plesiotype. Pleistocene. Melbourne, Fla. C. P. Single- 
ton leg., 1929. T. Barbour don. Distal part of left ulna. Wetmore, 
1931, p. 35. 

Hali^etus leucocephalus (Linnaeus) 

2232. Plesiotype. Pleistocene. Melbourne, Fla. C. P. Single- 
ton leg., 1928. T. Barbour don. Part of right metacarpus. Wet- 
more, 1931, p. 30. 

Jabiru mycteria (Lichtenstein) 

2225. Pleisotype. Pleistocene. Melbourne, Fla. C. P. Single- 
ton leg., 1929. T. Barbour don. Fragments of right tarsometatarsus, 
and three right and two left tibiotarsi. Wetmore, 1931, p. 17. 

Meleagris gallopavo Linnaeus 

2231. Plesiotype. Pleistocene. Melbourne, Fla. C. P. Single- 
ton leg., 1928. T. Barbour don. Parts of left humerus, right tibio- 
tarsus, and right tarsometatarsus. Wetmore, 1931, p. 33. 

Palaealectoris incertus Wetmore 

2190. Genoholotype. Lower Miocene. Agate Springs, Nebr. 
E. M. Schlaikjer leg., 1928. T. Barbour don. Left humerus (distal 
part of shaft missing). Wetmore, 1930, pp. 152-153, figs. 51-53. 



schevill: fossil types 101 



Palaeobonasa wyomingensis (Eastman) 

2221. Genoholot^-pe. Shufeldt, 1915, p. 633. 
See Gallinuloides wyomingensis Eastman. 

PaLAEOSPIZA BELLA Allen 

2222. GenoholotN-pe. Miocene. Florissant, Colo. S. H. Scudder 
leg., 1877. Exch. Boston Society of Natural History. Allen, 1878, 
pp. 443-445, pi. I, fig. 1; 1878a, pp. 381-384, fig. 1; 1878b, pp. 204- 
205, fig. 1 ; Wetmore, 1925, pp. 183-191, text figures 1, 2, and pis. I-IV. 

The original of Figure 2 in Allen's papers (cf. also Wetmore, 1925, 
p. 185) has not been found. - ^wW '9^^ ^ H-^--^- 

Palapteryx major Kneeland 

2220 (B.S.N.H. 10611). Cotype. Pleistocene. New Zealand. 
Coll. H. B. Cross. Exch. Boston Society of Natural History. Meta- 
tarsus. Kneeland, 1852, pp. 236-238; 1853, pp. 298-299. 

The femur, tibia, and phalanges mentioned by Kneeland cannot be 
located in the collections. 

Paractiornis perpusillus Wetmore 

2191. Genoholotype. Lower Miocene. Carnegie Hill, Agate, 
Nebr. E. M. Schlaikjer leg., 1929. Left tarsometatarsus. Wet- 
more, 1930, pp. 153-154, figs. 54-56 (the figures are x2, not natural 
size as noted). 

QUERQUEDULA FLORIDANA Shufeldt 

2227. Plesiotype. Pleistocene. Melbourne, Fla. C. P. Single- 
ton leg., 1928. T. Barbour don. Proximal part of right humerus. 
Wetmore, 1931, p. 22. 

Struthiolithus chersonensis Brandt 

2223 (formerly M.C.Z. 1597). Plesiotype. Pleistocene. Yao 
Kuan Chuang, ca. 50 miles southwesterly of Kalgan, China. Coll. 
W. P. Sprague, 1898. Eastman, 1898a, pp. 127-135, pi., figs. 1. 2. 



102 bulletin: museum of comparative zoology 



BIBLIOGR.\PHY 

Agassiz, Louis 

1832. Neues Jahrbuch fiir Mineralogie, etc., pp. 129-149. 

1833-1844. Recherches . . . poissons fossiles, 1-5, text and atlas. 4°. 

Neuchatel et Soleure, 1833-1843 (-1844). 
1835. Rev. crit. poiss. foss. figures dans I'lttiolitologia Veronese. 44 pp. 

8°. Neuchatel. 
1835a. Neues Jahrbuch fi'ir Mineralogie, etc., pp. 290-316. 
1856. Amer. Jour. Sci. (2), 21, No. 62, pp. 272-275. 
Allen, Joel Asaph 

1878. Bull. U. S. Geol. and Geogr. Surv. Terr., 4, No. 2, pp. 443-445. 
1878a. Amer. Jour. Sci. (3), 15, No. 89, pp. 381-384. 

1878b. Nature, 18, June 20, 1878, pp. 204 205. 
Barbour, Thomas 

1917. Proc. Biol. Soc. Washington, 30, pp. 97-103. 

1918. Bull. Mus. Comp. Zool., 61, No. 14, pp. 479-489. 

1919. Proc: Biol. Soc. Washington, 32, pp. 145-147. 
1926. Copeia, No. 151, pp. 109 111. 

Barbour, T., and Loveridge, A. 

1929. Bull. Mus. Comp. Zool., 69, No. 10, pp. 205-360. 
Barbour, T., and Stetson, H. C. 

1929. Bull. Mus. Comp. Zool., 69, No. 4, pp. 99 104. 
Blake, J. F. 

1876. Reptilia, in Tate, Ralph, and Blake, J. F., The Yorkshire Lias. 8°. 
London. 
Bronn, Heinrich Georg 

1828. Zeitschr. fiir Mineralogie, 1828, Bd. 1 (Taschenbuch fiir . . . 

Mineralogie, 22, Bd. 1), pp. 374 384. 
1830. Neues Jahrbuch fiir Mineralogie, etc., pp. 14-30. 
Camp, Charles L. 

1930. Mem. Univ. California, 10, \+ 161 pp., 6 pi. 
Claypole, E. W. 

1892. Amer. Geol., 10, No. 4, pp. 199 207. 
Cope, Edward Drinker 

1869. Proc. Acad. Nat. Sci. Philadelphia, June 1, 1869, p. 123. 
1869a. Nature, 1, November 25, 1869, pp. 121 122. 

1870. Trans. Amer. Philos. Soc. (n. 8.), 14, Pt. 1, pp. 1-252. 
1875. Rpt. U. S. Geol. Surv. Terr., 2 (Vert. Cret. Form. West). 

1879. Amer. Nat., 13, No. 5, p. 333. 

1884. Rpt. U. S. Geol. Surv. Terr., 3 (Vert. Tert. Form. West, Bk. 1). 
CusHMAN, Joseph A. 

1907. Proc. Boston Soc. Nat. Hist., 33, No. 6, pp. 249-275. 
Dawson, J. William 

1870. Quart. Jour. Geol. Soc. London, 26, pp. 166-167. 



schevill: fossil types 103 

1870a. Geol. Mag. (1), 7, pp. 87-88. 

1870b. Canad. Nat., (2), 5, pp. 98-99. 
Dean, Bashford 

1901. Mem. N. Y. Acad. Sci., 2, Pt. 3, ii, iii, pp. 101-123. 

1909. Mem. Amer. Mus. Nat. Hist., 9, Pt. 5, pp. 211-287. 
Eastman, Charles Rochester 

1896. Amer. Jour. Sci. (4), 2, No. 7, pp. 46-50. 

1897a. Ann. Rpt. Iowa Geol. Surv., 1896, pp. 108-116 (1897). 

1897b. Amer. Nat., 31, No. 366, pp. 493-499. 

1897c. Bull. Mus. Comp. Zool., 31, No. 2, pp. 19 44. 

1898. Amer. Nat., 32, Nos. 379, 380, 382, pp. 473-488, 545-560, 747-768: 
1898a. Bull. Mus. Comp. Zool., 32, No. 7, pp. 127-143. 

1899. 17th Ann. Rpt. N. Y. State Geol., 1897, pp. 317-327 (1899). 
1899a. Jour. Geol., 7, No. 5, pp. 489-493. 

1900. Geol. Mag., (4), 7, No. 2, pp. 54-58. 
1900a. Jour. Geol., 8, No. 1, pp. 32-41. 

1900b. Bull. Mus. Comp. Zool., 36, No. 3, pp. 67-75. 

1900c. Centralbl. flir Mineralogie, etc., 1900, No. 6, pp. 177-178. 

1901. Maryland Geol. Surv., Eocene, pp. 98-115. 
1902a. Geol. Mag. (4), 9, Nos. 4, 9, pp. 148-152, 388-391. 
1902b. Bull. Mus. Comp. Zool., 39, No. 3, pp. 55-97. 
1902c. Amer. Nat., 36, Nos. 428, 431, pp. 653-659, 849-854. 
1902d. Jour. Geol., 10, No. 5, pp. 535-541. 

1903. Mark Anniversary Volume, Art. 14, pp. 279-289. 
1903a. Mem. Mus. Comp. Zool., 26, No. 4, pp. 179-189. 
1903b. Bull. Mus. Comp. Zool., 39, No. 7, pp. 163-221. 

1904. Bull. Mus. Comp. Zool., 46, No. 1, pp. 1-35. 
1904a. Amer. Nat., 38, No. 453, pp. 669-672. 
1904b. Amer. Jour. Sci. (4), 18, No. 106, pp. 253-260. 

1905. New Jersey, Ann. Rpt. State Geol., 1904, pp. 27-102 (1905). 

1906. Bull. Mus. Comp. Zool., 50, No. 1, pp. 1-29. 
1906a. Bull. Mus. Comp. Zool., 50, No. 4, pp. 75-98. 
1907a. Bull. Mus. Comp. Zool., 50, No. 7, pp. 211-228. 
1907b. New York State Mus. Mem., 10. 

1908. Iowa Geol. Surv., 18, Ann. Rpt. 1907, pp. 29-386 (1908). 

1911. Connecticut State Geol. and Nat. Hist. Surv., Bull. 18. 
Hay, Oliver Perry 

1923. Pan-Amer. Geol., 39, No. 2, pp. 101-120. 

1923a. Pan-Amer. Geol., 40, No. 1, pp. 29-31. 
Heintz, Anatol 

1929. Skrifter om Svalbard og Ishavet Nr. 22, Oslo, 1929. 
HussAKOF, Louis 

1906. Mem. Amer. Mus. Nat. Hist., 9, Pt. 3, pp. 105-154. 

1908. Bull. Amer. Mus. Nat. Hist., 25, Pt. 1, pp. 1-103. 

1908a. Science (n. s.), 28, No. 714, September 4, 1908, pp. 311-313. 

1913. Bull. Amer. Mus. Nat. Hist., 32, Art. 11, pp. 245-250. 



104 bulletin: museum of comparative zoology 

HUSSAKOF, L., AND BrTANT, W. L. 

1919. Bull. Buffalo Soc. Nat. Sci., 12, 346 pp., 1918 (1919). 
Jackson, Charles T. 

1851. Rpt. Albert Coal Mine [New Brunswick], pp. 18-25. 8°. Boston. 

1852. Proc. Boston Soc. Nat. Hist., 4, pp. 138-143. 
Jager, G; 

1850. Nov. Act. Acad. Caes. Leopold-Carol. . . . , 22, Abt. 2 (Verb., 
14, Abt. 2), pp. 767-934. 

Janensch, W. 

1906. Beitr. Palaeont. Geol. Oesterr. . . . , 19, Heft 1, pp. 1-33. 
Kneeland, Samuel 

1852. Proc. Boston Soc. Nat. Hist., 4, pp. 236-239. 

1853. Proc. Boston Soc. Nat. Hist., 4, pp. 298-299. 
Knight, Wilbur C. 

1899. Jour. Geol., 7, No. 4, pp. 357-374. 
Lambe, Lawrence M. 

1909. Amer. Jour. Sci. (4), 28, No. 164, pp. 165-174. 

1910. Contrib. Canad. Paleont., 3 (4°), Pt. 5. 
Lambrecht, K. 

1916. Aquila, 23, pp. 196-307, 483-501. 

1921. Fossilium Catalogus (1, Animalia), 12, 104 pp. 
Leidy, Joseph 

1868. Proc. Acad. Nat. Sci. Philadelphia, pp. 177-178. 
Lucas, Frederic A. 

1900. Bull. Mus. Comp. ZooL, 36, No. 4, pp. 79-84. 
Lull, Richard Swann 

1920. Amer. Jour. Sci. (4), 50, No. 297, pp. 234-236. 
Massalongo, Abramo Bartolommeo 

1859. Specimen photographicum animalium . . . plantarumque fossiUum 
. . . (Saggio fotografico . . .). 4°. Verona. 

Merriam, John C. 

1902. Univ. Calif. Publ., Bull. Dept. Geol., 3, No. 4, pp. 63-108. 

1908. Mem. Univ. Cahfornia, 1, No. 1, pp. 1-156. 
von Meyer, Hermann 

1848. Neues Jahrbuch fiir Mineralogie, etc., pp. 781-784. 

1851. Neues Jahrbuch ftir Mineralogie, etc., pp. 75-81. 
1851a. Palaeont ographica, 2, Lief. 3, pp. 85-113. 

1855. Palaeontographica, 4, Lief. 3, pp. 80-83. 

1856. Palaeontographica, 6, Lief. 1, pp. 22-30. 

1860. Palaeontographica, 7, Lief. 3, pp. 123-182. 
1863. Palaeontographica, 10, Lief. 5, pp. 227-246. 

MooK, Charles C 

1932. Bull. Mus. Comp. Zool., 74. no. 2, pp. 19^1. 
Newberry, John Strong 

1883. New York Acad. Sci. Trans., 2, No. 8, pp. 144-147. 

1885. New York Acad. Sci. Trans., 5, No. 2, pp. 25-28. 



schevill: fossil types 105 

1888. Compte Rendu, 3°"® Congr. Geol. Internat. Berlin, 1885, pp. 1 1-14. 

1889. Monogr. U. S. Geol. Surv., 16 (Paleoz. Fish. N. Amer.). 

NiLSSON, S. 

1824. Kongl. Vet.-Acad. Handl., 1823, pp. 96-106 (1824). 
Parker, George H. 

1908. In Eastman, 1908 — Iowa Geol. Surv., 18, Ann. Rpt., 1907, p. 272 
(1908). 
Raymond, Percy E. 

1925. Amer. Jour. Sci. (5), 10, No. 60, pp. 551-555. 
St. John, Orestes H., and Worthen, A. H. 

1875. Palaeont. Illinois, 6, Pt. 2, pp. 245-488. 

1883. Palaeont. Illinois, 7, Pt. 2, pp. 52-264. 
Shufeldt, Robert W. 

1915. Jour. Geol., 23, No. 7, pp. 619-634. 
Stetson, Henry C. 

1927. Jour. Geol., 35, No. 3, pp. 247-263. 

1928a. Jour. Geol., 36, No. 5, pp. 458-470. 

1928b. Amer. Jour. Sci. (5), 16, No. 93, pp. 221-231. 

1930. Bull. Mus. Comp. ZooL, 71, No. 2, pp. 19-39. 

1931. Jour. Geol, 39, No. 2, pp. 141-154. 
Stock, Thomas 

1881. Ann. Mag. Nat. Hist. (5), 7, No. 42, pp. 490-492. 

1883. Ann. Mag. Nat. Hist. (5), 12, No. 69, pp. 177-190. 
Traquair, Ramsay H. 

1886. Ann. Mag. Nat. Hist. (5), 18, No. 108, pp. 493-496. 
Udden, J. A. 

1899. Jour. Geol., 7, No. 5, pp. 494-495. 

VOLTA, G. SeRAFINO (eT AL.) 

1796. Ittiolitologia Veronese, Pt. 2, ch. 4, pp. 263-277. f°. Verona, 
1796-1809. 
Wetmore, Alexander 

1925. Bull. Mus. Comp. Zool., 67, No. 2, pp. 183-193. 

1930. Condor, 32, No. 3, pp. 152-154. 

1931. Smithson. Misc. Coll., 85, No. 2. 
Woodward, Arthur Smith 

1889. Cat. Foss. Fishes Brit. Mus. (Nat. Hist.), Pt. 1. 

1890. Geol. Mag. (3), 7, No. 10, pp. 455-460. 

1891. Cat. Foss. Fishes Brit. Mus. (Nat. Hist.), Pt. 2. 
1901. Cat. Foss. Fishes Brit. Mus. (Nat. Hist.), Pt. 4. 

Woodworth, Jay B. 

1900. Bull. Geol. Soc. America, 11, pp. 449-454. 



FEB 20 1933 

Bulletin of the Museum of Comparative Zoology 

AT HARVARD COLLEGE 

Vol. LXXIV. No. 5 



BIRDS FROM NORTHWEST YUNNAN 



By James C. Greenway, Jr. 



CAMBRIDGE, MASS., U. S. A.: 
PRINTED FOR THE MUSEUM 
February, 1933 



PUBLICATIONS 
OF THE 

MUSEUM OF COMPARATIVE ZOOLOGY 
AT HARVAED COLLEGE 



There have been published of the Bulletin Vols. I to LXV, LXVII- 
LXXIV, of the Memoirs Vols. I to LI. 

The Bulletin and Memoirs are devoted to the publication of 
original work by the Officers of the Museum, of investigations carried 
on by students and others in the different Laboratories of Natural 
History, and of work by specialists based upon the Museum Collec- 
tions and Exploration. 

These publications are issued in numbers at irregular intervals. 
Each number of the Bulletin and of the Memoirs may be sold sepa- 
rately. A price list of the publications of the Museum will be sent on 
application to the Director of the Museum of Comparative Zoology, 
Cambridge, Massachusetts. 



Bulletin of the Museum of Comparative Zoology 

AT HARVARD COLLEGE 

Vol. LXXIV. No. 5 



BIRDS FROM NORTHWEST YUNNAN 



By James C. Greenway, Jr. 



CAMBRIDGE, MASS., U. S. A.: 

PRINTED FOR THE MUSEUIM 

February, 1933 



No. 5. — Birds from Northwest Yunnan 
By J.vmes C. Green way, Jr. 

Dr. Joseph F. Rock returned to the United States for a short visit 
after having made his fine collections for the United States National 
IMuseum. He returned to China to collect for the Museum of Com- 
parative Zoology. This is a report of the birds that have been sent 
to the Museum by him in the years 1931 and 1932. 

Upon his arrival at Yunnan-fu, capital of the westernmost province 
of China, in August, he found that the provincial government had 
long been at war with the neighboring province of Kwang-tung, that 
three thousand men had been sent to the war, and that in consequence 
the country had been left to the mercy of bandits. For this reason 
it was considered unsafe to proceed up country and a long delay fol- 
lowed. So it was not until January that he was able to lead his caravan 
of thirty-five mules over the long but to him familiar trail to the 
northwest. 

In about three weeks the expedition arrived at Likiang, seat of the 
district government, but they found that the Nah-si population 
were in the process of celebrating the festival of muan-mo (propitia- 
tion of heaven) and would do no work. After more delay the caravan 
with a guard of a few ill-armed soldiers, trailed off to the northwest, 
into the valley of the Mekong and later northward into the valley 
of the Salween and the headwaters of the Irrawaddy. 

The expedition was divided, and real collecting began in April and 
continued until March of the next year. I have listed the localities 
where Rock collected together with the meager information that he 
has furnished. 

A-dshwa, eastern slopes of the Likiang Snow Range, 14,500 ft., 
crags and cliffs, May, 1931. 

Champutong, border of Tibet and Yunnan, Salween Valley, eastern 
slopes of the Salween-Irrawaddy Divide, 9,000 to 10,000 ft., July. 

Chou-yu-gko, above Tao-mung-chung, Likiang District, 13,000 to 
15,000 ft., eastern slopes of the Yangtze-Mekong Divide, fir and rhodo- 
dendron forest, April, 1931. 

Gomba-la, Mt. (Kenichunpo), eastern slopes of the Salween- 
Irrawaddy Divide, 14,000 to 16,000 ft., alpine meadows and rocky 
crags and also fir and rhododendron forest, July, 1931. 

Gyi-na-loko, Mt., Likiang Snow Range, 10,000 to 15,000 ft., October 
or November, 1931. 



no 



bulletin: museum of comparative zoology 



Haba-ndsher-nvulu Snow Range, northwest of the Yangtze big 
bend, above 14,000 ft., fir and spruce forests with Arundinaria under- 
growth, January, 1932. 




Fig. 1 



Hsiao-Wei-hsi, north of Wei-hsi, banks of the Mekong River, Sep- 
tember, 1931. 

Kenichunpo, Mt. (see Mt. Gomba-la). 

La-shi-pa, west of Likiang, rice fields, September, 1931. 

Na-dza-gko, alpine meadow of, slopes of Mt. Satseto. 



GREEN way: birds FROM NORTHWEST YUNNAN 111 

Xv-lu-ko, foot of the Likiang Snow Range, 9,500 ft., July, 1931. 

Satseto, jMt., east slopes of the Likiang Snow Range, 12,000 to 
15,000 ft., limestone formation; spruce and fir forests mixed with 
Acer, Syringa, Euonymus, Arundinaria, October and November, 1931. 

Shi-ku, banks of the Yangtze River, 6,200 ft., willow and poplar 
groves, March, 1931. 

Shwe-men-kan, Mt., Haba-ndsher-nvulu Snow Range. 

Su-wa-tong, Tibet, 14,000 to 16,000 ft., upper slopes of Mt. Gomba- 
la or Keniclumpo, July, 1931. 

Tao-mung-chung, southwest of Lu-tien (Li-tien), Likiang District, 
10,000 to 12,000 ft., southern slopes of the Yangtze-Mekong Watershed 
(Li-ti-ping), spruce and fir forests, April and May, 1931. 

Tse-chung, Mts., IMekong Valley, eastern slopes of the Mekong- 
Salween Divide, 8,000 to 11,000 ft., forests, August and September. 

Tung-la, Mts. (To-la in Tibetan), above Ho-fu-ping, western slopes 
of the Yangtze-Mekong Divide (Pe-ma-shan), 12,000 to 14,000 ft., 
fir and rhododendron forests, August, 1931. 

Wei-hsi, west of, 9,000 ft., pine forests, June, 1931. 

AVei-hsi, Mts., west of, east of the Mekong River, 10,000 to 11,000 
ft., spruce and hemlock forests with canebrake undergrowth, June, 
1931. 

Wei-hsi, banks of the Wei-hsi River, 7,000 ft., June, 1931. 

Yun-nan-yi, plain of, east of Talifu, 5,000 ft., February, 1931. 

Yung-ning, Lake, two days northeast of the Yangtze big bend, 
spruce forests with canebrake undergrowth, 10,000 ft., February, 
1932. 

The collection contains over 1,800 skins, comprising 216 forms. 
It is a well made collection, without overburdening series of common 
things and with its share of rarities. Cholcites xanihorhynchis has 
been added to the Yunnan list and three forms have been described. 
They are Erytkrina edwardsii ruhicunda, Erythrina vmacca ruhidior 
and Ithaginis cruentus holoptilus. 

Thanks are due to the United States National Museum for the 
loan of needed material and to Joseph H. Riley, N. B. Kinnear of 
the British Museum of Natural History and B. Stegmann of the 
Academy of Science, Leningrad, for advice. 

It would be impossible adequately to thank Outram Bangs and 
James L. Peters. Since the report was done under their constant 
kindly supervision, their work merges subtly into mine, so that it 
would be impossible to say where one began and the other ended. 
During the period in which the collection was being identified Outram 



112 bulletin: museum of comparative zoology 

Bangs died. I cannot express the value of his influence and help not 
only to this little paper but also to my education. 

Phasianus colchicus elegans Elliot 

Ann. Mag. Nat. Hist., 6, 1870, p. 312 (Yung-ling Mountains, W. Sechuan). 

A single male was taken on the slopes of ]\It. Gyi-na-loko in April; 
three males at Tao-mung-chung in April or May and a female from 
the mountains of Tung-la (To-la) in August. 

Chrysolophus amherstiae (Leadbeater) 

Phasianus amherstiae Leadb., Trans. Linn. Soc. London, 16, 1828, p. 129, 
pi. 15 (Mountains of Cochin China). 

Five mature males, two immature and two females were taken at 
Tao-mung-chung in April or May; an immature male at Mt. Satseto 
in December; a female at Mt. Gyi-na-loko in October or November, 
and another female at an unnamed locality in the Likiang Range in 
March, and two males at A-dshwa in May. 

Pucrasia meyeri Madrasz 

Ibis, 1886, p. 145 (Central Tibet). 

Single males come from Tao-mung-chung (April or May), Mt. 
Gyi-na-loko (October or November), and three males and a female 
from the Likiang Snow Range (February). 

Tetraophasis szechenyii Madrasz 

Zeitschr. Ges. Orn., 2, 1885, p. 50, pi. 2 (East Tibet). 

Four males and six females were taken on the Likiang Snow Range 
in February and three males and a single female at Mt. Gyi-na-loko 
on the Likiang Range in October or November. 

Crossoptilon crossoptilon (Hodgson) 

Phasianus crossoptilon Hodgson, Journ. Asiat. Soc. Bengal, 7, 1838, p. 864, pi. 
46 (Tibet). 

A male and three females were taken at Mt. Satseto in February. 

Verreaux mentions three characters to distinguish his dromjnii 
(tibetanus). They are (1) grayish-white remiges instead of brown, 
(2) outer rectrices which lack white spots and (3) the median rectrices 
narrower and without the green-gold tinge. 



GREEN way: birds FROM NORTHWEST YUNNAN 113 

In a series of three pairs taken by Zappey in southwest Sechuan 
in July and August the primaries are distinctly brownish while those 
of our present series are grayish. In males of both series the white 
outer webs of the outer primaries are apparent. I conclude that the 
primaries of summer birds are faded to brown and that the white 
on the outer primaries is a variable sex character. 

I have not seen any specimens from central Tibet so that I cannot 
tell about the relative amount of white on the outer rectrices nor the 
relative width of the middle feathers. In the series of the Museum of 
Comparative Zoology a single male from Sechuan has the outer tail 
feathers edged with gray as has the single male from the Likiang 
Range, Yunnan. In view of these facts it would seem doubtful that 
drouynii is a tenable form. 

Tragopan temmincki (Gray) 
Satyra temmincki Gray, in Hardwicke's III. Ind. Zool., 1, 1830-32, pi. 1 (China). 

Rock sent a single mature male from Shwe-men-kan (January); 
two females from Mt. Gyi-na-loko, one taken in November and the 
other in April; an immature male and two females from Mt. Satseto 
(December), and an immature male and two females from Ghou-yu- 
gko (April). 

Ithaginis cruentus kuseri Beebe 

Zoologica, 1, 1912, p. 190 (Yunnan). 

A single male specimen was taken at Mt. Gomba-la or Kenichunpo 
in southeast Tibet, elevation 14,000 to 16,000 ft., in June. 

Ithaginis cruentus holoptilus subsp. nov. 

Type. — No. 160, 786 Museum of Comparative Zoology from Chou-yu- 
gko, above Tao-mung-chung, Likiang District; east slopes of the 
Yangtze-Mekong Divide, 13,000 to 15,000 ft., collected in April, 1931. 

Characters. Intermediate between Ithaxiinis c. kvseri Beebe and 
Ithaginis c. rocki Riley. On comparison with rocki the lores will be 
found to be black and red instead of black, and the same difference 
will be found in the line above the eye; the sides of the neck are white 
narrowly edged with gray rather than gray with narrow white shaft 
lines; sides of the upper breast are red and green rather than red mixed 
with buffy, or buffy. The chief character upon which I have separated 
this subspecies is the texture of the feathers of the crest, which are 
normal rather than decomposed as in clarkei, rocki and kuseri. 



114 bulletin: museum of comparative zoology 

Rock has sent three males and three specimens sexed as females. 
I think, however, that two of these are immature males since they 
have red feathers appearing on the neck and breast; all of these were 
taken at Chou-yu-gko, Likiang District. 

The range of these subspecies is, of course, very imperfectly known. 
As far as is known, however, this new form has an intermediate range. 
Kuseri, according to Rothschild, has been taken from Tseku, where its 
range borders on that of rocki, the Salween Valley, Tengueh and the 
Shweli-Salween Divide; rocki to the north in the mountains of Ho-fu- 
ping, and clarkci in the Likiang Mountains to the westward. 

As well as a single specimen of Ithaginis c. clarkei Roths., taken at 
the Likiang mountains in May, I have examined two males and a 
female of clarkei from the same locality, kindly loaned by the United 
States National Museum, as well as two males a.nd a female oilthaginis 
c. rocki from the Ho-fu-ping mountains. 

Ithaginis cruentus clarkei Rothschild 
Bull. B. O. C, 40, 1920, p. 67 (Likiang Range, Yunnan). 

Ten males were taken at ^It. Satseto in January and February and 
one at Gyi-na-loko in October or November, two from Shwe-men-kan 
(January) and one at A-dshwa in May. Six females come from Mt. 
Satseto (January or February), one from Mt. Gyi-na-loko (October 
or November) and one from Shwe-men-kan (January). 

Amaurornis fuscus erythrothorax (Temminck & Schlegel) 
Gallinula erythrothorax T. & S., Siebold's Faun. Jap., Aves, 1849, p. 121 (Japan). 

Seven males, six females and one unsexed specimen were taken on 
the banks of the Wei-hsi River in June. 

Males measure wing: 98-107 and females 106-108. 

Sphenocercus sphenurus yunnanensis La Touche 
Bull. B. O. C, 42, 1921, p. 13 (Lotukow, Yunnan). 

Four males come from the mountains of Tung-la (To-la) (August). 

COLUMBA LEUCONOTA GR.-VDARIA Hartert 
Novit. ZooL, 23, 1916, p. 85 (Szetchuan). 

Rock has sent a single specimen without data. 



GREENWAY: birds from northwest YUNNAN 115 









5 



o 



«^ 



^^ «J Ih flj 

-ij a; 

g S ^ i 

3 t^ ii • s 



TJ 






















(D 


-C 






whit 
nes 


-«-j 
■? 


^ 

3 


CO 


o t; 


"C 


_r;, 


O X 


1^ 

T3 -k^ 






o 




Short; 
fea 














-C "C O 



03 -^ =r j^ 
cc' C ^ 



o3 



2 ^ ^ I --«--= t1 -^ « 7 

c:S^^ ;^«Ph 



o 



c 



^^■^ 






5^S « -giffl S ^^^ 









"oi 


o 




O/ 


a 


-i«i 


t>^ 


E 


03 




o 


ffl 


o 


0) 




^ 


-a 




OQ 







Woo 
B: c3 C3 

2 5 S 



-o 


"c3 
43 


43 


^ 


o 




r^ 


p. 

a 


03 
G3 




c3 


o 
o 


-i-i 




4<! 


(U 


13 


o 


-o 


& 


'-I3 


^ 


bio 




(3 
o 


3 

c 


>> 


-C 


05 


o3 






a 





>^ t« 43 



0} 



OS 






00 


feathe 
osed 


huffy 
nes 


huffy 
nes 


O 00 


ery long; 
decompi 


Gray with 
shaft li 


Gray with 
shaft li 


1—1 


> 









0) 



^ « 



« o o _ 

O 42 03 2.72 

J <3 W U t» 





03 


O 


03 

o 


oo 


42 




kH 


o 


03 


C» 


a 




& 




3 



116 bulletin: museum of comparative zoology 

Streptopelia orientalis orientalis (Latham) 
Columba orientalis Latham, Ind. Orn., 2, 1790, p. 606 (China). 

Two males and a female come from Tao-mung-ehung (April or 
May), a single male from west of Wei-hsi (June) and a male without 
data. 

Eulabeia indica (Latham) 

Anas indica Latham, Ind. Orn., 2, 1790, p. 839 (India in winter and Tibet). 
Anser indicus, Stuart Baker, Faim. Brit. Ind., ed. 2, 4, 1929, p. 405. 

A female and an unsexed specimen were shot on the plain of Yun- 
nan-yi, east of Talifu in February. 

Ardea cinerea jouyi Clark 
Proc. U. S. Nat. Mus., 32, 1907, p. 468 (Seoul, Corea). 

One female was taken at La-shi-pa, west of Likiang in rice fields 
in September. This is clearly a white necked form although the skin 
is in rather poor condition. 

BuLBULCUS ibis coromandus (Boddaert) 
Cancroma coromanda Boddaert, Tabl. Pi. Enl., 1783, p. 54 (Coromandel) . 

A male in breeding plumage comes from Tao-mung-chung in April 
or May. 

MiLVUS LiNEATus (Gray) 

Haliaetus lineatus Gray, 111. Ind. Zool., 1, 1832, p. 1, pi. 18 (China). 

One unsexed specimen, a female, comes from the Likiang Snow 
Range (February). 

Accipiter virgatus affinis Hodgson 

Bengal Sport. Mag., New Ser., 8, 1836, p. 179 (Nepal). 

Rock has sent what by its ^'ery dark coloration appears to be an 
immature female from the Likiang Snow Range (April). 

BUTEO BURMANICUS BURMANICUS Hume 

Stray Feath., 3, 1875, p. 30 (Thayetmyo, Burma). 

Two males come from the Likiang Snow Range (eastern slopes) and 
were taken in February. 



GREEN way: birds FROM NORTHWEST YUNNAN 117 

Spizaetus nipalensis nipalensis (Hodgson) 

Nizaetus nipalensis Hodgson, Journ. Asiat. Soc. Bengal, 5, 1836, p. 229, pi. 7 
(Nepal). 

A single specimen sexed as a female was taken at Mt. Gyi-na-loko 
in April. The wing measures 455 mm. It is an immature bird. 

Aquila nipalensis nipalensis Hodgson 

Asiat. Researches, 18, pt. 2, 1833, p. 13, pi. 1 (Nepal). 

One female comes from the foot of the Likiang Snow Range, 9,400 
ft. (IMarch). 

Falco tinnunculus japonensis Ticehurst 

Bull. B. O. C, 50, 1929, p. 10. 

Falco tinnunculus japonicus Temminck and Schlegel, in Siebold's Faun. Jap., 
Aves, 1844, p. 2, pi. 1 (Japan). [Name preoccupied.] 

An immature male and female come from Mt. Gyi-na-loko (October 
or November). 

Gl.^ucidium cuculoides whiteleyi (Blyth) 

Athene whiteleyi Blyth, Ibis, 1867, p. 313 (China). 

A single female was taken at Shi-ku, on the banks of the Yangtze 
in ]\Iarch. This bird appears to be somewhat darker than a long 
series from Sechuan, Hupeh and Fukien in the Museum of Compara- 
tive Zoology. 

Strix nivicola (Blyth) 

Syrnium nivicolum Blyth, Journ. Asiat. Soc. Bengal, 14, 1845, p. 185 (Himalayas). 

A single male specimen was taken at Chou-yu-gko in April. 

Bangs, ^ La Touche,- Rothschild ^ and Stuart Baker '* seem to be 
agreed that S. aluco harterti La Touche and Strix niripetcns Riley 
are synonyms of nivicola: that the wide individual variations of these 
owls cannot be grouped in any way in order to show that there is a 
race of the Indian bird in China. I have foUoAved Bangs in keeping 
7iivicola specifically distinct until there is more evidence that will 
point to its affinities with aluco. 

1 Bull. Mus. Comp. Zool. Harvard, 70, No. 4, 1930, p. 197. 

2 Birds of East. China, 2. pt. 2, 1932, p. 109. 
5 Novit. Zool., 33, 1926, p. 2.33. 

'Faun. Brit. Ind., 4, 1927, p. 399. 



118 bulletin: museum of comparative zoology 

Chalcites xanthorhynchus (Horsfall) 

Cuculus xanthorhynchus Horsfall, Trans. Linn. Soc. 13, pt. 1, 1S21, p. 179 
(Java). 
A single female was taken north of Wei-hsi on the Mekong in Sep- 
tember. This is the first record of this bird for Yunnan. It is, of course, 
surprising that it has never been taken before, but since it is known 
to breed in Assam it is not incredible. 

Cuculus canorus bakeri Hartert 
Vog. Pal. Faun., 2, 1912, p. 948 (Shillong, Khasia Hills). 

Rock has sent a series of eight mature birds, a male from Tao-mung- 
chung (April or May), two males, three females and two unsexed 
specimens from Chou-yu-gko (April). 

Two birds just from the nest were taken at Champutong in July. 
These specimens have the exposed portion of the wing barred with 
brown. 

Cuculus optatus optatus (Gould) 

Proc. Zool. Soc. London, 1845, p. 18 (Port Essington, Australia). 

A single male in juvenal plumage was shot in the mountains of 
Tse-chung in August or September. 

Cuculus poliocephalus poliocephalus Latham 

Ind. Orn., 1, 1790, p. 214 (Srinagar). 

Two males come from Chou-yu-gko (April), a female from Champu- 
tong (July) and an unsexed specimen from the mountains of Tse- 
chung (August or September). 

Megalaema virens virens (Boddaert) 
Bucco idrens Bodd., Tabl. PI. Enl., 1783, p. 53 (China). 

An unsexed immature specimen was taken at Champutong in July. 

PicuMNUs innominatus chinensis (Hargitt) 
Vivia chinensis Harg., Ibis, 1881, p. 228, pi. 7 (May-chee, China). 

A single male was taken in canebrake at Tao-mung-chung in April 
or May. 



GREEN way: birds FROM NORTHWEST YUNNAN 119 

Macropicus forresti (Rothschild) 

Dryocopm forresti Roths., Bull. B. O. C, 43, 1922, p. 9 (Mekong Valley, 
Yunnan). 

A female was taken at Tao-mung-chung in April or May and 
another at Chou-yu-gko in April. An unsexed specimen comes from 
the mountains of Tung-la (To-la) in August. 

PicoiDES FUNEBRis Vcrreaux 
Nouv. Arch. Mus. Paris, Bull., 6, 1870, p. 33 (Mountains of Chinese Tibet). 

An unsexed specimen comes from the mountains of Tung-la (To-la) 
(August), a male and an unsexed specimen from Mt. Gyi-na-loko 
(October or November), a female from Mt. Satseto (February) and 
two males and a female from Shwe-men-kan (January). 

Dryobates NANUS OMISSUS (Rothschild) 

Dryobates pygmaeus omissus Roths., Bull. B. O. C, 43, 1922, p. 10 (Likiang 
Range). 

A pair comes from Shwe-men-kan (January) and a single female 
from Mt. Satseto (February). 

These specimens are larger than Dryobates obscurus La Touche 
(wing 104-106-103) and are somewhat less heavily striped and more 
buffy. This accords very well with the original description. Roths- 
child has emended his own description ^ and finds that omissus differs 
from obscurus only in its larger size. 

Dryobates hyperythrus hyperythrus (Vigors) 

Picus hyperythrus Vigors, Proc. Zool. Soc. London, pt. 1, 1831, p. 23 (Himalayan 
Mountains). 

A male, a female and an unsexed specimen come from Tao-mung- 
chung (April or May), a pair from Chou-yu-gko (April), a female 
from Mt. Gyi-na-loko (October or November) and a female from Mt. 
Satseto (February). 

I have followed Rothschild in referring this series to the Indian 
form. There seems to be no tinge of the dull brown with its lack of 
red that is characteristic of subrvfinus, nor is the back whiter than 
in Indian birds. 

1 Novit. Zool., 33, 1926, p. 238. 



120 bulletin: museum of comparative zoology 



Dryobates darjellensis desmursi (Verreaux) 

Picus desmursi Verr., Nouv. Arch. Mus. Paris, Bull., 6, 1870, p. 33 (Mountains 
Chinese Tibet). 

* 

A female comes from Tao-miing-ehung (April or May), a female, 
an unsexed specimen and an immature bird, sexed as a female, from 
the mountains of Tung-la (To-la) (August). 



Dryobates major stresemanni Rensch 

Zool. Erg. W. Stotz. Exp., in Abh. und Ber. Mus. Dresden, 16, 1924, no. 2, 
pt. 3, p. 38 (Tsalila). 

Two males and a female were taken at Tao-mung-chung in April 
or May, a single female at Chou-yu-gko in April, two females at Mt. 
Gyi-na-loko (October or November) and a male and two females at 
Mt. Satseto in February. 



Picus canus sordidior (Rippon) 
Gecinus sordidior Rippon, Bull. B. O. C, 19, 1906, p. 32 (W. Yunnan). 

Two males were taken at Tao-mung-chung in April or May, a 
single female west of Wei-hsi in June, an unsexed specimen in the 
mountains of Tung-la (To-la) in August, and a female at Mt. Gyi-na- 
loko in October or November. 

Psittacula schisticeps finschi (Hume) 
Palaeornis finschi Hume, Stray Feath., 2, 1874, p. 509 (Kollidoo). 

A male was taken at Shi-ku on the banks of the Yangtze in March 
and an immature male at Nv-lu-ko at the foot of the Likiang Snow 
Range in July. 

Psittacula derbyana (Fraser) 

Palaeornis derhyanus Fraser, Proc. Zool. Soc. London, 1850, p. 245, pi. 25 
(No locality). 

A male comes from Na-dza-gko, eastern slopes of Mt. Gyi-na-loko 
(April) and an unsexed specimen at Mt. Gji-na-loko in October or 
November. 



GREENWAY: birds from northwest YUNNAN 121 

Spelaeornis rocki Riley 

Proc. Biol. Soc. Wash., 42, 1929, p. 214 (Mountains of Ho-fu-ping, Mekong 

Valley, Yunnan). 

As Riley remarks, it is a strange thing to find a distinct species so 
close to the type locality of Spelaeornis souliei Oustalet. I have 
compared the specimens at hand carefully with the figure of the type 
of souliei} As Riley found in comparing the type of rocki, so I find 
that the birds that Rock has sent are lighter above and the black 
apical spots are more conspicuous, that the flanks are lighter and the 
black apical spots of that region are smaller, and that the white of 
the throat extends to the jugulum. Of these characters I consider that 
the last only is of any value. Due to the fact that the spring 
birds are in worn and faded plumage (the ones at hand are distinctly 
so) and that all the birds in the Paris Museum were mounted and 
exposed to the light until very recent times and that consequently 
most of them are "foxed" badly, it can be readily understood that 
the colors of the new skins might very easily be quite different from 
the type and the differences be just such as are described above. 
However, the white of the throat does extend to the jugulum and for 
that reason I recognize rocki. 

A female and an unsexed specimen come from Tao-mung-chung 
and a female and an immature bird from Chou-yu-gko (April). 

Nannus troglodytes talifuensis (Sharpe) 
Anorihura talifuensis Sharpe, Bull. B. O. C, 13, 1902, p. 11 (Gyi-dzin-shan). 

Two males, a female and an immature specimen were taken at 
Tao-mung-chung (April or May), two pairs at Mt. Gyi-na-loko 
(October or November), a male and two females at Mt. Satseto (De- 
cember) and two males, a female and one unsexed specimen from the 
Likiang Snow Range (F'ebruary). 

In comparison with a series of birds taken in Sechuan in winter 
these birds are appreciably darker. There is, however, a good deal 
of individual variation. 

A curious fact came to my notice in examining the series in the 
Museum of Comparative Zoology. It is that there are two specimens 
from Choni (Nos. 135,791, 135,792) which are indistinguishable 
from the Yunnan bird. They are quite mature and were taken in 

1 Novit. Zool., 17, 1910, pi. 7, fig. 1. 



122 bulletin: museum of comparative zoology 

winter. From another area they might be thought to belong to an 
intermediate form, but Choni should harbor an intermediate between 
the light Nannus t. szcischuanen^is Hartert and the even lighter 
NannU'S t. idiu^ (Richmond) rather than a very dark form. 

Prunella immaculata (Hodgson) 
Accentor immacidata Hodgson, Proc. Zool. Soc. London, 13, 1845, p. 34 (Nepal). 

A single male was taken on Mt. Gyi-na-loko in October or November 
and five more at different levels of the west slopes of Mt. Satseto in 
December, January and February. 

Prunella strophiata multistriata (David) 
Accentor multistriatus David, Ann. Mag. Nat. Hist., 7, 1871, p. 256 (Moupin). 

Three males, three females and two unsexed specimens were taken 
on the east slopes of the Likiang Snow Range in November and 
December; two males, three females on the west slopes of Mt. Satseto 
in January and February; a male and an unsexed specimen at Tao- 
mung-chung in April or May; a single female at Chou-yu-gko in May 
and three unsexed specimens at Su-wa-tong, Tibet, in July. 

Prunella collaris ripponi Hartert 
Vog. Pal. Faun., 1, 1910, p. 766 (Gyi-dzin-shan). 

Of a series of six birds, five come from the Likiang Snow Range, a 
male was taken at Mt. Gyi-na-loko in October or November, a female 
in February at Mt. Satseto and a male and two females in March. 
A single male was taken at Su-wa-tong, Tibet, in July. 

Prunella collaris berezowskii (Serebrowski) 

Laiscopus collaris berezoivskii Serebrowski, C. R. Acad. Sci. Leningrad, 1927, A, 
p. 325 (Lun-ngan-fu, Sechuan). 

Serebrowski described this bird as being lighter than Prunella c. 
nipaleruns (Blyth) and darker than PruncUa c. tihetamis Bianchi. 
I add that it is like Prunella c. ripponi Hartert but larger. The wing 
measures 103 mm. in this specimen. It is also more dusky on the 
flanks. 

Rock has sent a single male from Mt. Satseto (March). 



GREENWAY: birds from northwest YUNNAN 123 

Enicurus maculatus guttatus Gould 
Proc. Zool. Soc. London, 1865, p. 664 (Sikkim). 

A single immature specimen of this bird was taken at Champutong 
in July. 

Enicurus leschenaulti sinensis Gould 

Proc. Zool. Soc. London, 1866, p. 665 (Shanghai). 

Two males were taken at Chou-yu-gko (April) and an immature 
specimen sexed as a female from the banks of the Wei-hsi River at 
9,000 ft. in June. 

Calliope tschebaiewi Przewalski 
Mongol I. Stran. Tang., 2, 1876, p. 4-4. 

A single male comes from Tao-mung-chung (April or May). 

Larvivora brunnea Hodgson 
Journ. Asiat. Soc. Bengal, 6, 1837, p. 102 (Nepal). 

A single malp was taken at Tao-mung-chung in April or May. 

Phoenicurus schisticeps (Gray) 

Rutidlla schisticeps Gray, Cat. Mamm. Birds Nepal Coll. Hodgson, 1846, p. 
69 (Nepal). 

In a series of twenty-two birds the wings measure: 81-86 for the males 
and 78-80 for the females. Five males were taken at Mt. Gyi-na-loko 
(October or November), two males and a single female at Shwe-men- 
kan (January), two males and seven females at Mt. Satseto at the 
end of January and the beginning of February; seven females at the 
same place in March. 

Phoenicurus frontalis frontalis Vigors 
Proc. Comm. Zool. Soc. London, 1832, p. 172 (Himalaya). 

Three males and four females were taken at Mt. Gyi-na-loko in 
October, November, a pair on Mt. Satseto in December and a single 
male in March; two males and a single female come from Tao-mung- 
chung (April or May) . 

The autumn and winter males of this series have the feathers of the 
back tipped with reddish brown and the females are dark brown on the 
back rather than gray as are the spring and summer specimens. 



124 bulletin: museum of compakative zoology 

Phoenicurus aurorea aurorea (Pallas) 

Motacilla aurorea Pallas, Reis. Russ. Reichs., 3, 1776, p. 695 (Selenka, Lake 
Baikal). 

Two males were taken at Tao-mung-chung (April or May) and a 
female at Chou-yu-gko (April). They are no darker than specimens 
taken by Rock on the Kansu border of Sechuan in 1928. I have, there- 
fore, followed Stuart Baker in considering the Yunnan bird to be 
identical with the Indian and northern Chinese birds. 

Phoenicurus hodgsoni (Moore) 

Ruticilla hodgsoni Horsfall & Moore, Cat. Birds Mus. E. Ind. Co., 1854, p. 308 
(Nepal). 

A male comes from Shwe-men-kan (January), a female from Mt. 
Satseto (February) and a female from Tao-mung-chung (April or 

May). 

Rhyacornis fuliginosa fuliginosa (Vigors) 

Phoenicura fuliginosa Vigors, Proc. Zool. Soc. London, 1831, p. 35 (No locality). 

Six males and four females come from Tao-mung-chung, seven 
males and two females from Chou-yu-gko (April), two males from the 
mountains west of Wei-hsi (June) and one immature male from 
Su-wa-tong (July). 

Chaimarrhornis leucocephala (Vigors) 

Phoenicura leucocephala Vigors, Proc. Zool. Soc. London, 1831, p. 35 (Hima- 
layas). 

Rock has sent a male from Su-wa-tang, Tibet, (July) and two 
females from Tao-mung-chung (April or May). The male was molting. 

Tarsiger chrysaeus vetellinus Stresemann 
Journ. f. Orn., 71, 1923, p. 365 (Washan). 

A series of fifteen mature specimens proves this race to be lighter 
than the Indian bird which is noticeably more orange. 

Six males and nine females come from Su-wa-tong, Tibet, (July), 
as do eleven immature specimens. A single female was taken at Mt. 
Gyi-na-loko (October or November). An immature bird taken at 
the mountains of Tung-la in August seems to have the flanks yellow; 
birds taken a month earlier do not. 



GREENWAY: birds from northwest YUNNAN 125 

MusciSYLViA leucura Hodgson 

Proc. Zool. Soc. London, 1845, p. 27 (Nepal). 

A single immature specimen (unsexed) was taken at Champutong 
in July. 

Rhodophila ferrea haringtoni (Hartert) 

Oreiciclaferreaharingtoni HarteTtyYog.Fal.Faun., l,p. 711 (Lien-kiang bei Fu- 
tschau, China). 

Rock has sent a series of thirteen birds; a single male from Tao- 
mung-chung (April or ^lay), seven males, a female and two immature 
specimens from the mountains west of Wei-hsi (June) and a male 
from the mountains of Tse-chung (August or September). 

Ianthia indica yunnanensis (Rothschild) 

Tarsiger indicus yunnanensis Rothschild, Bull. B. O. C, 43, 1922, p. 10 
(Likiang Range). 

A male and three females were taken at Tao-mung-chung (April or 
May) and a single female at Chou-yu-gko in April. 

Ianthia rufilata practica Bangs and Phillips 
Bull. Mus. Comp. Zool., 58, 1914, p. 292 (Loukouchai). 

Of a series of twenty-one of these birds, a male and two females 
were taken at Tao-mung-chung (April or May), a single female at 
Chou-yu-gko (April), and three males and ten females from Mt. 
Gyi-na-loko (October or November). Three immature specimens 
were taken at the mountains of Tung-la and at Su-wa-tong (July). 

Riley has remarked that males exhibit two color phases, one cyanite 
blue above and the other marine Vjlue. This I find to be true of my 
specimens, except that the blue is rather closer to alizarine than to 
navy. This difference in coloration is shown by two specimens that 
were taken at the same locality at the same time. 

Saxicola torquata przewalskii (Pleske) 

Pratincola maura var. przewalskii Pleske, Wis. Res. Przewalski's Reis., Vog. 1, 
1889, p. 46 (Kansu). 

Two males, a female and an immature specimen are from Tao- 
mung-chung (April or May), a single male from the Likiang Snow 
Range (March). 



126 bulletin: museum of comparative zoology 

These birds are noticeably larger and darker than Saxicola t. indica 
(Blyth) and Saxicola t. stcjnegeri (Parrot). 

Myiophoneus coeruleus (Seopoli) 
Gracula coerulea Scop., Del. Flor. Faun. Insubr., 2, 1786, p. 88 (China). 

A female comes from the mountains of Tung-la (To-la) (August) 
and a male from Mt. Gyi-na-loko (October or November). 

MoNTicoLA RUFivENTRis (Jardine and Selby) 

Petrocincla rufiventris J. &. S., Illust. Orn., 3, pi. 129, 1828 (Himalayas). 
Turdus erythrogaster Vigors, Proc. Zool. Soc. London, 1831, p. 171. 
Monticola erythrogastra Baker, Faun. Brit. Ind., 2, p. 170. 
Monticola erythrogaster La Touche, Handb. Birds East. China, 1925, p. 123. 
Monticola rufiventer sinensis Meinertzhagen, Bull. B. O. C, 47, 1927, p. 148. 

An unsexed immature specimen and an immature female were 
taken in the mountains of Tung-la (To-la), above Ho-fu-ping, on the 
Mekong side of the Yangtze-Mekong Divide, 12,000 to 14,000 ft. in 
August. 

These two specimens differ somewhat from the description in 
La Touche's " Handbook of the Birds of Eastern China," which is 
the most complete and satisfactory, and I therefore describe them 
here. 

Feathers of the head and upper back are black with a large white 
spot; on the lower rump they are barred olive and black, and the 
upper tail coverts are barred bay and black. The ear coverts are 
black, some of the feathers having white shafts. Feathers of the 
throat have the white spot very large. On the breast the feathers 
have large yellow spots, and on the flanks these spots are brownish 
yellow. 

A nestling from the La Touche collection, taken in Kuatun, N. W. 
Fukien on May 22, agrees very closely with the present series. The 
spots on the feathers are more uniformly yellow, however, and this 
character conforms more nearly to the general impression of yellow- 
ness that descriptions give. I believe that the difference is probably 
due to age, partly of the wild specimen when it was shot and partly 
to the "foxing" or fading toward brownish that skins undergo in 
museums. 

Female specimens of nifivrnfris from China in this museum are not 
"darker or more slate colored above" nor have they "the ground 



GREENWAY: birds from northwest YUNNAN 127 

color of the feathers of the under parts more black brown, giving a 
darker appearance," ^ than specimens from Sikkim. I cannot find 
that the Chinese bird differs in any way from the Indian. 

TuRDUs RUBROCANUS GOULDi (Verreaux) 

Merula gouldi Verr., Nouv. Arch. Mus. Hist. Nat. Paris, Bull., 6, 1871, p. 34 
(western Sechuan). 

This series is made up of three males from Tao-mung-chung, taken 
in April or May, two males and two females from Chou-yu-gko, taken 
in April, and a pair, a single female and an unsexed specimen from 
the mountains of Tung-la (To-la), taken in August, and four males 
from Mt. Satseto (October or November). 

This bird appears to be confined to high altitudes. I cannot find 
that it has ever been taken below 10,000 ft. 

The series is remarkably uniform. There are none of the individual 
differences reported to occur. The females are lighter than the males. 

TuRDus NAUMANNi EUNOMUS Tcmminck 
PI. Col., 1830, p. 514 (Japan). • . 

Of nine specimens in the series there are four males, one from the 
banks of the Yangtze at Shi-ku, two from the eastern slopes of Mt. 
Satseto (December), and a fourth from the eastern slopes of the 
Likiang Snow Range (February). Four females come from Tao-mung- 
chung (April or May) and a single female from Shwe-men-kan (Jan- 
uarv). 

Dr. B. Stegmann of the Academy of Sciences, Leningrad, writes 
that the breeding area of Turdus n. cunomus Temm. lies to the north- 
ward of Turdus n. naumanni Temm. The breeding area of cunomus 
is Siberia, from the southern edge of the tundra and the lower course 
of the Jenissei, south to the northern Baikal Range and east to 
Bering Sea. Turdus n. naumanni Temm. breeds from the middle 
reaches of the Jenissei River over the northern Baikal Range and south 
Jakutia to the Stanoi Range. The breeding areas of these two birds 
do not overlap. 

Dr. Stegmann believes that the differences in coloration are great 
enough so that these birds might be considered as distinct species. 
However, he also says that their habits and voice are very similar 
and that they may with reason be thought to be subspecies. 

Since their breeding ranges do not overlap, their habits are similar 
and they are so similar in appearance, I prefer to list them as subspecies. 

> vide Meinertzhagen Bull. B. O. C, 47, 1927, p. 148. 



128 bulletin: museum of comparative zoology 

TuRDUS NAUMANNi NAUMANNi Temminck 
Man. d'Orn., 1, 1820, p. 170 (Eastern Europe). 

A male comes from Tao-mung-ehung and a female from the Likiang 
Snow Range, east slopes of Mt. Satseto. The former was taken in 
April or May and the latter in March. Two unsexed specimens were 
shot at the foot of the Likiang Snow Range in March. 

The male is darker than the female, having the feathers of the upper 
breast marked with a chestnut subterminal bar while the female has 
the bars amber. 

TuRDUS OBSCURUS Gmelin 
Syst., Nat., 1, 1789, p. 816 (Lake Baikal). 

Three females were taken at Mt. Gyi-na-loko in October or No- 
vember. 

TuRDUS MUPINENSIS CONQUISTUS Bangs 
BuU. Am. Mus. Nat. Hist., 46, 1921, p. 591 (Likiang Range). 

Two males were taken at Tao-mung-chung in April or May. 

TuRDUS RUFICOLLIS RUFICOLLIS Pallas 

Reise Russ. Reichs, 3, 1776, p. 694 ("In Summis ingis Dauuriae"). 

Only one specimen, a male, was collected at Tao-mung-chung in 
April or May. 

TuRDUS MOLLISSIMUS MOLLISSIMUS Blyth 
Joiu-n. Asiat. Soc. Bengal, 11, 1842, p. 188 (Darjeeling). 

Rock has sent two males from the east slopes of Mt. Gyi-na-loko 
taken in October or November. These birds have the terminal black 
bars of the underparts not crescentic like the typical bird but rather 
drawn straight across the feather. There are also fewer of these black 
bars so that the underparts appear lighter and the belly pure white. 
The color of the back is a light greenish brown, not the dark olive 
brown of the Indian bird. 

Of a series of ten skins kindly loaned to me by the United States 
National Museum, nine are identical with the Indian bird, the tenth, 
however, is much like the birds that Rock has just sent. It was shot 
at Mt. Omei, Sechuan in December. 



GREENWAY: birds from northwest YUNNAN 129 

Breeding specimens of mollissiinus have been taken in northwest 
Yunnan and southeast Sechuan. It does not seem very probable 
that there is a geographical representative isolated in this region, 
mountainous and given to curious faimal variations as it is. 

TuRDus DAUMA socius (Thayer and Bangs) 

Oreocincla dauma soda Thayer & Bangs, Mem. Mus. Comp. Zool., 11, No. 4, 
Aves, 1912, p. 174 (Tatsienlu). 

A single male comes from Tao-mung-chung (April or May). 

POMATORHINUS RUFICOLLIS SIMILIS Rothschild 

Novit. Zool., 33, 1926, p. 261 (Hills near Tengyueh). 

A single female taken in the mountains of Tung-la (To-la) in 
August has the outer webs of the primaries brown instead of grayish 
green and a browner appearance than the other specimens in the series. 
Only the base of the upper mandible is black, whereas two-thirds of 
the upper mandible is black in the other birds of this series. The wing 
measures 81 mm., the tail 100 mm., the exposed culmen 18 mm. 

There are eleven specimens in the series, three males and two 
females from Tao-mung-chung (April or May), two males and three 
unsexed specimens from Mt. Satseto (December). 

POMATORHINUS MACCLELLANDI DEDEKENSI Oustalet 

Ann. Sci. Nat. Zool., ser. 7, 12, 1892, pp. 276, 304 (Southern Shensi). 

A male and two unsexed specimens have been sent from Tao-mung- 
chung (April or May), a pair from Chou-yu-gko (April) and three 
pairs from the western slopes of Mt. Satseto in December, January or 
Februarv. 

Pomatorhinus m. gravivox David, the eastern form, seems to be 
browner and less green, season for season, than Pomatorhinus m. 
dedekensi Oust, from western Sechuan and northern Yunnan. It is 
also smaller, the wing measuring 90-95 mm. as against 98 and over. 
Pomatorhinus m. odicus Bangs and Phillips, which is the southern 
Yunnan form, is smaller still, but resembles dedekensi very closely 
in other respects. 

I have examined six specimens of gravivox from Hupeh taken in 
May, June, October, November, January and February, two speci- 
mens of dedekensi from the Likiang range (June) and two from south- 



130 bulletin: museum of comparative zoology 

western Sechuan (May and January), as well as a very long series of 
odicus from the vicinity of Mongtz. There is a great deal of individual 
variation in the size of the bill as well as the colors in this group. The 
color of the back seems to have "foxed" or grown browner in the 
older specimens. 

Garrulax albogularis albogularis (Gould) 
lanthocincla albogularis Gould, Proc. Zool. Soc. London, 1835, p. 187 (Nepal). 

Six males and eight females have been sent from Tao-mung-chung 
and Chou-yu-gko in April and May. These birds are slightly darker 
than specimens of albogularis from northern India in the Museum of 
Comparative Zoology. 

Garrulax subunicolor griseata (Rothschild) 

lanthocincla subunicolor griseata Roths., Novit. Zool., 28, 1921, p. 33 (Schweli- 
Salween divide). 

To Rothschild's description I add that the specimens at hand have 
the feathers of the cheeks and ear coverts with whitish shafts, con-, 
trasting with the head and neck. 

Two pairs, a male and a female have the primaries molting. They 
were taken at Su-wa-tong, Tibet, in July. 

Garrulax affinis oustaleti (Hartert) 
lanthocincla affinis oustaleti Hart., Vog. Pal. Faun., 1, 1909, p. 633 (Tsekou). 

Two males and a female were taken at Tao-mung-chung in April 
or May and a pair at Chou-yu-gko in April, eight males on the Likiang 
Snow Range in October, November and December and two unsexed 
specimens from the east slopes of the same range in February, seven 
males, six females and six unsexed immature specimens from Su-wa- 
tong, Tibet, (July) and a single male from Su-wa-tong in January. 
A nestling was taken at the mountains of Tung-la (To-la) in August. 

The nestling agrees perfectly with the description that Rothschild 
gives. ^ "Plumage differs from the adults by the brown, not black, 
crown, the absence of the gray patch at the side of the neck, and the 
uniform brown of upper and underside." 

I Novit. Zool.. 33, 1926, p. 264. 



GREENWAY: birds from northwest YUNNAN 131 

Garrulax elliotii elliotii (Verreaux) 

Trochalopteron elliotii Verreaux, Nouv. Arch. Mus. Paris, Bull., 6, 1870, p. 36 
(Mountains of Chinese Tibet). 

Two males, five females and two unsexed specimens were taken 
at Tao-mung-chung in April or May; two males and one female at 
Chou-yu-gko in April; a male and two unsexed specimens at the 
mountains of Tung-la (To-la) in August ; a single female at Champu- 
tong (Chamutang) in July; an unsexed specimen on the eastern slopes 
of Mt. Satseto in March, and five males, three females and one unsexed 
specimen at jNIt. Gyi-na-loko in October or November. This series 
of twenty-two specimens was taken at all seasons of the year in the 
Likiang District. 

After a careful comparison of these birds and a good series in the 
Museum of Comparative Zoology, it is apparent that the silvery 
white tips to the feathers, particularly of the throat, are lost by a 
process of wearing in the spring and summer. For this reason there is 
little doubt that Garrulax bonvaloti, which Oustalet described "with 
many reservations" from Chatou, Tibet, as being without the little 
white marks on the head, cheeks, neck and breast, is a synonym of 
Garrulax c. clliofii (Verr.). Riley ^ and Berlioz ^ are agreed about this. 

Winter birds are likewise somewhat darker and grayer than sum- 
mer birds. Although I have no specimens identified as Garrulax c 
yunnannise Rippon, described from the Yangtze River, Likiang 
Range, Yunnan, as darker and altogether grayer than elliotii, I am 
inclined to believe that Riley is right that this character is due to 
seasonal variation and that yunnanense cannot be maintained as a 
valid race. 

Furthermore these birds conform very closely to the description of 
Garrulax e. prjevalski which Menzbier described ^ from Kansu. As 
Bangs and Peters have said, "* the gray instead of greenish tail feathers 
seems to be the only distinguishable character by which prjevalski 
can be separated from elliotii. The present series shows gradation 
from gray to green. 

Garrulax cineracea styani Oustalet 

Bull. Mus. Paris, 6, 1898, p. 226 (Ta-tsien-lou). 

Two specimens of this bird, both males, were taken in the moun- 
tains of Tse-chung in August or September, 1931. 

1 Proc. U. S. Nat. Mus., 80, 1931, p. 37. 

2 Rev. Hist. Nat. 1. 2, 3, 1930. 

3 Ibis, 1887, p. 300. 

* Bull. Mus. Comp. Zool., 68, 1928, p. 341. 



132 bulletin: museum of comparative zoology 

These two specimens have the cheeks white, the superciliary line 
olive gray and the ear coverts a faded color approaching buffy yellow. 
Some of the feathers of this region are white at the base and buffy 
yellow only at the tip. I place them w^ith styani after Rothschild, ' 
rather than La Touche,- because of the plate of cinereiccps with Styan's 
description,^ which shows the chestnut ear coverts and superciliary 
line quite plainly. 

There are, however, in the collections of the Museum of Comparative 
Zoology, two specimens with the olive gray superciliary line and the 
faded buffy yellow ear coverts, which obviously are styani, but which 
were taken at Szemao, south Yunnan, on February 28, far south of 
the range assigned to them by Rothschild. Bangs in an unpub- 
lished paper which he has kindly allowed me to see, has suggested 
that these birds migrated south to Szemao in winter. 

Garrulax bieti (Oustalet) 

lanthodnda bieti Oust., Bull. Mus. Paris, 3, 1897, p. 163 (Tsekou, upper 
Mekong, Yunnan). 

Rock has sent a male and three females from Tao-mung-chung, 
taken in April or May. There is also an unsexed specimen from Chou- 
yu-gko, above Tao-mung-chung. 

There is a great deal of individual variation in the color of the head 
and the extent of the subterminal black bars of the feathers of the 
side of the neck and the flanks. I am inclined to agree with Riley * 
that these are age characters. 

A specimen sexed as a female, with the fluffy plumage and relatively 
pointed remiges of immaturity, has the feathers of the head medal 
bronze and the subterminal black bars of the sides of the neck and the 
flanks are lacking. 

Garrulax lanceolatus lanceolatus (Verreaux) 

PterorkiniLS lanceolatus Verr., Nouv. Arch. Mus. Paris, Bull., 6, 1871, p. 36 
(Mountains of Chinese Tibet). 

Four males, a female and an unsexed specimen were shot at Tao- 
mung-chung (April or May). 

Even in this small series the character of the color of the mousta- 
chial line does not run true. In two specimens it is almost auburn 

1 Rothschild, Novit. Zool., 33, 1926, p. 264. 

^ Birds of Eastern China, 1, p. 60. 

5 Ibis, 1887, p. 167, pi. 6. 

«Proc. U. S. Nat. Mus., 80, 1931, p. 34. 



GREENWAY: birds from northwest YUNNAN 133 

with black tips to the feathers, while in the others it ranges from 
darker brown to black. 

I consider Garrula.v bonraloti Oust, and Garrulax yunnanensis 
Rippon to be synonyms of kmccolafus. 

Garrulax ocellatus similis (Rothschild) 

lanthocincla ocellata sitnilis Roths., Novit. Zool., 28, 1921, p. 34 (Shweli-Salween 
Divide). 

This bird is apparently an intermediate form between Garrulax 
ocellatus ocellatus (\ ig.) and Garrulax ocellatus artemisiae (David and 
Oustalet). It resembles ocellatus rather than artemisiae in that the 
color of the ear coverts is rather brown than black or gray; this brown 
is, however, much lighter than the single specimen of ocellatus in the 
Museum of Comparative Zoology. The superciliary line is as light 
as in artemisiae. 

The character which Lord Rothschild mentions for artemisiae in 
his description oi similis ' ". . . feathers of the breast and foreneck hav- 
ing large black markings," does not seem to be borne out in the small 
series in the Museum of Comparative Zoology. There is, however, 
a female of artemesiae which has the color of the ear coverts brownish 
gray and this would lead one to believe that there must be a more 
distinctly intermediary form which is now furnished by the specimen 
in question. 

Apparently forms like the present one have never been taken. 
Rock collected this one from Chou-yu-gko in April. 

Garrulax maximus maximus (Verreaux) 

Pterorhinus maximus Verreaux, Nouv. Arch. Mus. Paris, Bull., 6, 1870, p. 36, 
pi, 3 (No locality). 

Four males and two females come from Mt. Gyi-na-loko (October 
or November) and a female from Shwe-men-kan (January) and two 
males from the Likiang Mountains (west slopes) taken in February, 
1932. 

Garrulax maximus khamensis (Serebrowski) 

lanthocincla maxima khamensis Serebrowski, C. R. Ac. Sci. de I'U. R. S. S., 
Leningrad, 1927, p. 325 (Kham country, E. Tibet). 

Unfortunately Rock has sent only a single specimen of this bird 
from the mountains of Tung-la (To-la) (August). It is apparently an 

I Novit. Zool., 28, 1921, p. 34. 



134 bulletin: museum of comparative zoology 

immature bird, but in comparison with other immature birds from 
Sechuan it is noticeably smaller. This is the only character that holds. 
The paler underside and the narrower dark bars of the chest seem to 
be age characters as Berlioz has remarked. ^ Its small size has led me 
to place it with the Kham bird but not without a good deal of misgiving. 

Garrulax sannio sannio Swinhoe 
Garrvlax sannio Swinh., Ibis, 1867, p. 403 (Amoy, Fukien). 

It is apparent from the long series in the Museum of Comparative 
Zoology, taken at all seasons of the year from Fukien to southern and 
western Yunnan, that this bird is darker in winter than in summer 
and that this change is due largely to the wearing of the feathers. 
Birds killed in the province of Hupeh in November and December are 
as dark as those taken at Mongtz, southern Yunnan, at the same 
season. Everywhere they all become lighter as the season goes on. 
There is a molting bird shot at Mongtz in June in the collections of 
this museum. 

La Touched says "the birds at Mongtz laid in April, and the 
young were about the first week in May." There is no molting bird 
of September or October in the collection. The plumage of the Nov- 
ember birds is quite fresh, however. The literature on the birds of 
China is sadly lacking in information on the subject of migrations. 

It seems to me probable that Garrulax albospccularis (Godwin- 
Austin) is Garrulus sannio sonnio Swinh. in worn plumage and for 
this reason I follow Rothschild and Riley in relegating this name to 
synonomy. 

Garrulax poecilorhyncha ricinus (Riley) 

Dryonastes berthemyi ricinus Riley, Proc. Biol. Soc. Wash., 43, 1930, p. 80 
(Ndamucho, Yunnan). 

In this series are two males from Chou-yu-gko (April or ]May). 

This series in comparison with four specimens from the La Touche 
collection which were taken from the type locality, Kuatun, Fukien, 
in April and May are more olivaceous and less rufous on the head 
and back and the cinnamon buff of the breast is lighter. The forehead 
is, however, not more strongly and extensively tinged with tawny. 

1 Rev. d'Hist. Nat., no. 1, 1930, p. 14. 
 Birds of Eastern China, p. 57. 



GREENWAY: birds from northwest YUNNAN 135 

I add to Riley's description that the terminal white bands of the 
outer rectrices are slightly broader and the webs of the outer rectrices 
are more dusky rather than brownish. 

Since the Yunnan bird has the same color pattern as jioecilorhyn- 
chus of Formosa, only differing from it in the lighter color, I prefer 
to consider the group as a " formenkreis " and I concur with LaTouche' 
and Stresemann - in doing so. 

LlOPARUS CHRYSOTIS FORRESTI (Rothschild) 

Fulvetla chrysotis forresti Roths., Bull. B. O. C, 46, 1926, p. 64 (Shweli-Salween 
Divide). 

There can be no doubt that these birds are forresti. In comparison 
with birds collected by Weigold at Kwantsien,of which Rothschild speaks 
in the original description, and of which I have two specimens at hand, 
they are much deeper orange below, the primaries are edged with 
orange and the throat is flecked with white. 

Rock has sent a series of six, four males, one female and one unsexed 
specimen from Tao-mung-chung (April or ]Ma^'). 

I prefer to consider swinhoei and forresti as well as chrysotis as geo- 
graphical representatives of the same group. Because of the very 
striking difference in color, the shorter hind claw and the smaller bill, 
I have followed Stuart Baker in retaining the genus Lioparus rather 
than lump these birds with the genus Fulvetta. 

FULVETTA RUFICAPILLA SORDIDIOR (Rippon) 
Proparus sordidior Rippon, Bull. B. O. C, 13, 1903, p. 60 (W. Yunnan). 
Five males and four females were taken at Mt. Satseto in December. 

Fulvetta striaticollis striaticollis (Verreaux) 

Siva striaticollis J. Verreaux, Nouv. Arch. Mus. Paris, Bull., 6, 1870, p. 38 
(Moupin). 

Three specimens, two males and a female come from Shwe-men-kan 
(January). 

These specimens are somewhat lighter below than the specimens 
from Sechuan in the collections of the Museum of Comparative Zoology. 
There are intermediates, however, and since the birds that Rock has 

2 Abh. u. Ber. d. Mus. f. Ti'erk. u. Volkerk. zu Dresden, 16, No. 2, 1923, p. 23. 



136 bulletin: museum of comparative zoology 

sent are winter specimens and those in the Museum of Comparative 
Zoology were taken in summer, I am inclined to believe that lightness 
is a seasonal character. 

FULVETTA CINEREICEPS YUNNANENSIS (Rothschild) 

Proparus striaticollis yunnanensis Roths., Bull. B. 0. C, 43, 1922, p. 11 (Mekong- 

Salween Div.). 
Fulvetta imperata Riley, Proc. Biol. Soc. Wash., 43, 1930, p. 123. 

Because there have been several errors made in the identification 
of this bird in the past, I have had a good deal of difficulty in diagnosing 
it. In the first place Rothschild erred in placing it with the striaticollis 
"formenkreis." It undoubtedly belongs with the cinereiceps group. 
Secondly, this mistake confused Riley when he found this bird in the 
collection sent to the United States National Museum by Rock. I 
have the type of his Fulvetta imperata at hand and it is identically the 
same as my birds, which Kinnear has seen and identified as Fulvetta s. 
yunnanensis Roths., after comparing them with specimens in the British 
Museum. 

The Chinese forms of the genus may be identified by the following 
key: 

A. Flanks and belly alike 

a. Ear coverts black vinipcdus bieti 

b. Ear coverts striped striaticollis 

B. Flanks and belly different 

a. Distinct white ring around the eye ruficapilla 

b. No white ring around the eye 

b'. Three innermost primaries edged with brown cinereiceps 
b". Three innermost primaries all black fvcata 

The cinereiceps group may be diagnosed as follows: 

A. Head gray; no black stripe 

a. Back distinctly reddish brown cinereiceps 

h. Back lightly washed with red fcssa 

B. Head brown; indistinct black lines guttaticollis 

C. Head sooty browTi; distinct black lines yumianctms 

tonkincnsis 

There is no character that will serve to distinguish yunnanensis 
and Fulvetta c. tonkincncsis Delacour, which may conveniently be put 
into a key, though when specimens of each are seen together the differ- 



GREENWAY: birds from northwest YUNNAN 137 

ences are marked. Fuhrfia c. tonlxincnsis is much darker. The brown 
of the flanks and rump is deeper and redder and the crown is a darker 
gray. 

I have never seen a specimen of Fuketta ruficapilla manipurensis 
Grant and I am not sure that it belongs to the ruficapilla group. Kinnear 
has wTitten to me as follows: "... Fulretta r. manipurensis . . . has the 
head intermediate in color between yunnancnsis and sordidior, the back 
tinged with the color of the head and the rump richer — an ochraceous 
brown — below, the thighs and abdomen are similar to the rump and 
the breast is also tinged with the same color as the back." 

As far as is known the ranges of the Chinese forms of Fulvetta may 
be outlined as follows: 

Fulvetta vinipectus bieti Oust.; from the Mekong Valley in southeast 
Tibet, east to Tatsienlu where it breeds in late May (Weigold), south to 
Tseku and the Shweli-Salween Divide. 

Fulvetta s. striaticolli^ (Verr.); Upper Tebbuland, Kansu, south to 
Tatsienlu, where it breeds in July (Weigold). 

Fulvetta r. ruficapilla (Verr.); Tsin-ling mountains, northern Sechuan, 
south to Kwan and Tankwan, southern Sechuan. 

Fidvetta ruficapilla sordidior Roths. ; mountains of Ho-fu-ping, Yangtze- 
Mekong Divide, east to the Likiang Range and south to the Tengueh 
District. 

Fulvetta c. cinereiceps (Verr.); Western Sechuan, where it has been 
taken in spring, summer and early autumn, east to Hsien-shansien, 
Hupeh, where Zappey took it in December, and Ichang, Hupeh, from 
which locality an autumn bird was described by Styan. It is recorded 
as " resident " in Hupeh by Gee, MofYet and Wilder.^ 

Fulvetta c. fessa Bangs and Peters; Ha Tebbuland and Choni, Kansu. 

Fulvetta c. guttaticollis La Touche; Fukien (resident), Kwantung 
(winter). 

Fulvetta c. yunnanensis Roths.; Mekong ^"alley, INIekong Salween 
Divide; Yangtze-Mekong Divide. 

Fulvetta fucata Styan; Hupeh; Ho-chaping (in the mountains north 
of the Yangtze) (April) and Ichang (December). 

It is not yet possible to say whether Fulvetta c. cinereiceps and Fulvetta 
fucata have overlapping breeding areas, nor do we know whether Fulvetta 
c. ymmanemis Roths, and Fulvetta r. sordidior Roths, nest in the same 
locality somewhere in their ranges. 

Rock has sent a pair and a single unsexed specimen from Tao-mung- 
chung (April or May) and a male from Chou-yu-gko, taken in April. 

1 Pekin Soc. Nat. Hist., Bull., 1, 1926, p. 177. 



138 bulletin: museum of comparative zoology 

FULVETTA VINIPECTUS BIETI (Oustalet) 

Aldppe bieti Oust., Ann. Sci. Nat., 12, 1892, p. 283, pi. 9, fig. 2 (Ta-tsien-lou). 

This good series consists of two males, two females and two unsexed 
specimens from Tao-mung-chung (April and May); four males from 
Chou-yu-gko (April), two females and two unsexed specimens from 
the Mountains of Tung-la (August), a male from the Likiang Snow 
Range; a male and an unsexed specimen from Su-wa-tong, Tibet, in 
July. In December Rock got four pairs, a female and an unsexed speci- 
men on Mt. Satseto, and in February a male and two females. 

MOUPINIA POECILOTIS SORDIDIOR Rothschild 

Novit. Zool., 28, 1921, p. 36 (Likiang Range). 

Rock has sent four males, one from the Likiang Snow Range 
(March), tw^o from the Shwe-men-kan (January) and one from 
Likiang Range (February). 

SCHOENIPARUS DUBIUS GENESTIERI (Oustalet) 

Alcippe genestieri Oust., Bull. Mus. Hist. Nat. Paris, 3, 1897, p. 210 (Tsekou). 

A male and one unsexed specimen were taken at Tao-mung-chung 
(April or May) and another male was taken in the mountains west of 
Wei-hsi in June. Four males and seven females come from Mt. 
Satseto. One male and five females were taken in February, the 
others in December. 

Pseudominla castaneiceps castaneiceps (Hodgson) 
Minla castaneiceps Hodgs., Ind. Rev., 1838, p. 38 (Nepal). 

One unsexed specimen comes from Champutong (Chamutang) 
(July). 

This appears to be the extreme northern and eastern extent of the 
range. Stuart Baker ^ gives the range east to the Shan States and the 
hills of central and east Burma. Rothschild ^ reports that Forrest 
was the only collector to send this bird from Yunnan and he only sent 
eight from the Tengueh District and the Schweli-Salween Divide. 
It is curious that La Touche has not recorded the bird from southern 
Yunnan, since Delacour ^ found it quite common at Chapa, near Lao- 

> Faun. Brit. Ind., p. 289. 

2 Novit. Zool., 23. 1926. p. 270. 

3 Ois. de rindo-Chine, p. 308. 



GREENWAY: birds from northwest YUNNAN 139 

Kay, in northern Tonkin, although he has not found it elsewhere, nor 
did Van Tyne ^ take it in upper Laos. In the present state of knowl- 
edge, therefore, its range is discontinuous and it is isolated at Chapa, 
250 miles from the border of the Shan States. 

Heteroxenicus cruralis formaster Thayer and Bangs 
Mem. Mus. Comp. ZooL, 40, 1912, p. 169 (Washan, W. Sechuan). 

Anyone attempting to identify birds of this group will find that a 
good deal of confusion has existed. There have been three forms de- 
scribed : 

1. Heteroxenicus c. cruralis Blyth, ranging from Simla eastward, 
both north and south of the Brahmaputra, through Assam, the Chin 
and Cachin Hills of northern Burma to S. W. Yunnan (two specimens 
of this form from Mongtz are in the Museum of Comparative Zoology). 

2. Heteroxenicus c. laurentii La Touche, which is only known from 
the type, which comes from Mongtz and is in the Museum of Com- 
parative Zoology. Since this bird was shot in October it would seem 
probable that it was on migration in southern Yunnan. It is indis- 
tinguishable from 

3. Heteroxenicus c. formaster Thayer and Bangs, the type of which 
comes from Washan in western Sechuan, and is recorded by Riley ^ 
from the Likiang Range. The wing of the type specimen measures 
73 mm. as does the wing of the type of Heteroxenicus c. laureniii La 
Touche. 

Rock has sent a male from Tao-mung-chung, taken in April or May 
and a pair from Chou-yu-gko, taken in April. The males are at once 
recognizable as the Chinese bird because of the stout bill and I have 
placed them with formaster in the belief that laurentii is a synonym 
of that form, La Touche having somehow, unaccountably, overlooked 
it. The female which Rock has sent resembles the Indian bird a little 
more in color than it does the Chinese, but because this shade of 
olivaceous brown is variable and subject to fading and the size of the 
bird (wing 71 as against 65 for females of the Indian bird in the 
Museum of Comparative Zoology), I have placed it ^-ith. formaster . 

There is one more difficulty. Rothschild has recorded^ Brachyp- 
teryx (Heteroxenicus) c. cruralis Blyth from the Likiang Range. I 
think it probable that since Rothschild had never seen specimens of 
the large billed form from China, the specimens that he has recorded 
as cruralis are really formaster. 

» Field Mus. Nat. Hist. Zool. ser., 18, 1931. 
2 Proc. U. S. Nat. Mus., 80, 1931, p. 51. 
' Novit. Zool., 33, 1926, p. 271. 



140 bulletin: museum of comparative zoology 

Leioptila desgodinsi desgodinsi (David and Oustalet) 

Sibia desgodinsi Dav. & Oust., Bull. Soc. Philom. Paris, 1, 1877, p. 139 (Yer- 
ka-lo). 

Two pairs come from the mountains of Tung-la (To-la) (July), a 
male from Champutong (July), a male, a female and an unsexed speci- 
men from Tao-mung-chung (April or May), two males from Chou-yu- 
gko (April), a male and two females from Mt. Gyi-na-loko (October 
or November) and two males and three females from Mt. Satseto 
(December and February). 

Leioptila pulchella coeruleotincta Rothschild 

Novit. Zool., 28, 1921, p. 38 (Shweli-Salween Divide). 

Two males and three females were taken at Su-wa-tong, Tibet, in 
July. 

Stachyridopsis ruficeps bhamoensis (Harington) 
Ann. Mag. Nat. Hist., ser. 8, 2, 1908, p. 245 (Bhamo). 

Three specimens, two males and a female, were taken on the 
mountains of Tse-chung in August or September and a pair at Cham- 
putong (Chamutang) in July. 

These birds are clearly the breeding birds of the region. They are 
molting. They resemble very closely Stachyridopsis r. bangsi La 
Touche but they differ in having the black lines of the throat much 
more clearly defined and much larger. Their upper plumage is rather 
faded as in other summer birds in the series of the Museum of Com- 
parative Zoology. 

Siva strigula yunnanensis Rothschild 
Novit. Zool., 28, 1921, p. 40 (Likiang mountains). 

This is a good series of twenty-five specimens, a male and two 
females from Tao-mung-chung (April or May), two pairs from the 
mountains of Tung-la (To-la) (August), eight males, seven females 
and an unsexed specimen from ]Mt. Satseto (December) and a single 
female from Shwe-men-kan (January). 

The series from Tao-mung-chung and the mountains of Tung-la, 
taken in spring and summer are all in the gray plumage, which accord- 
ing to Rothschild ^ and Riley ^ is characteristic of birds at this season 

> Novit. Zool., 28, 1921, p. 40. 

= Proc. U. S. Nat. Mus., 70, 1926, p. 30. 



GREENWAY: birds from northwest YUNNAN 141 

of the year. Winter birds taken at Mt. Satseto and Shwe-men-kan are 
in the brownish plumage. 

The irides of these specimens are all recorded as "bright red or 
crimson" while the eye of Siva s. castaneicauda Hume is recorded by 
Stuart Baker ^ as deep brown, as it is on the label of a specimen taken 
at Simla in the collections of the Museum of Comparative Zoology. 



YUHINA GULARIS GRISEOTINCTA Rothschild 
Novit. Zool., 28, 1921, p. 42 (Shweli-Salween Divide). 

Three specimens, all males, two from the Mountains of Tung-la or 
To-la, taken in August, and one from Su-wa-tong, Tibet, taken in 
July, are distinctly darker than the specimens of Yuhina g. yamipiensis 
Sharpe from Sechuan. Although they are summer birds and in worn 
plumage, the crest is a deeper shade of brown, contrasting with the 
back; the throat and chest are washed with a more deep vinous tinge 
and the belly is a deeper buff. A pair from Shwe-men-kan, taken in 
January and three males and a female from Mt. Satseto are still 
darker. 

The birds of this series have stouter bills than the Sechuan specimens 
in the collections of the Museum of Comparative Zoology. They measure 
18-19 mm., while four specimens of yangpiensis measure 16-17 mm. 

They compare closely with a single specimen of Yuhina g. gularis 
Hodgs., but since this bird was taken in 1872 it has probably "foxed" 
with time. 

As Rothschild has remarked, this is a bird of high altitudes. 



YumNA DI.VDEMATA AMPELINA RippOH 
Bull. B. O. C, 11, 1900, p. 12 (Warar-Bum east of Bhamo). 

Rock has sent sixteen specimens. Five males and four females were 
taken from Tao-mung-chung (April or May); three males and two 
females from Chou-yu-gko (April); a male from the Likiang Snow 
Range, eastern slopes of ]Mt. Satseto (March) and an immature male 
from the Mountains of Tung-la (August). Six males, four females and 
one unsexed specimen were taken at Mt. Satseto in October, November, 
December and February. 

> Faun. Brit. Ind., 1, 1922, p. 314. 



142 bulletin: museum of comparative zoology 

YuHiNA occipitalis obscurior Rothschild 
Novit. Zool., 28, 1921, p. 42 (Likiang Range). 

Rock has sent ten males and two females from Chou-yu-gko (April), 
seven pairs and six unsexed specimens from the mountains of Tung-la 
(To-la) (August), seven males and three females from Tao-mung-chung 
(April or May), one male and two females from Su-wa-tong, Tibet, (July) 
and one male from Mt. Satseto (December). 

YUHINA NIGRIMENTUM INTERMEDIA Rothschild 

BuU. B. 0. C, 43, 1922, p. 11 (Mekong VaUey). 

Twelve specimens are much darker below than a series of Yuhina n. 
pallida La Touche from Kuatun, Fukien. They have grayer throats. 

In this series there are six males, three females and three unsexed 
specimens all from Champutong (Chamutang), taken in July. 

Erpornis xantholeuca griseiloris Stresemann 
Joum. f. Orn., 71, 1923, p. 364 (Siuhang, Kwangtung). 

Having only a single specimen of the Indian bird for comparison, 
and that an old one, I follow Stresemann in separating the Yunnan 
bird, although, as he says, the character of the color of the head appears 
to be variable. Furthermore, I do not believe that the character of the 
greenness or yellowness of the back is stable, and I separate this bird 
with reluctance. 

Rock has sent one specimen from Champutong (Chamutang), taken 
in July. 

Myzornis pyrrhoura Hodgson 
Journ. Asiat. Soc. Bengal, 12, 1843, p. 984 (Nepal). 

This bird is apparently rare as far east as Yunnan and confined to 
high altitudes as it is in Nepal and Sikkim.^ Rock has never taken 
it before and Forrest has only sent six to Rothschild from the Schweli- 
Salween Divide and four from the Mekong-Salween Divide. In his 
report Rothschild says " that Oustalet reported a specimen sent to hun 
by the Reverend Father Soulie from Tsekou in 1900. 

Five males, two females and two unsexed specimens (Jul\') from 
Su-wa-tong. 

1 Faun. Brit. Ind., 1, 1922, p. 344. 

2 Novit. Zool., 33, 1926, p. 278. 



GREENWAY: birds from northwest YUNNAN 143 

Pteruthius aerulatus ricketti O.-Grant 
Bull. B. O. C, 14, 1904, p. 92 (Kuatun, Fukien). 

Two pairs were taken at Tao-mung-cliung (April or May), a pair 
at Mt. Satseto (December) and a male at Shwe-men-kan (January). 

Pteruthius xanthochloris pallidus (David) 

Allotrius xanthochloris var. pallidus David, Nouv. Arch. Mus. Paris, BuU., 7, 
1871, p. 14 (frontiers of Kookonor). 

A pair come from Tao-mung-chung (April or May), an unsexed 
specimen from the mountains of Tse-chung (August or September) and 
three pairs and a female from Mt. Satseto (December). 

SuYA sp. 

There is an immature bird in the collection which I cannot identify 
surely. It is perhaps Stiya parviwstris La Touche. 

Alctppornis nipalensis yunn'anensis (Harington) 

Alcippe fratercula yunnanensis Harington, Bull. B. O. C, 33, 1913, p. 63 (Gyi- 
dzin-shan, east of TaUfu). 

In comparison with two specimens of Alcippornis n. fratercula Rippon 
in the Museum of Comparative Zoology, I find that the heads of the 
series of seventeen birds in question are lighter gray, that the blackish 
superciliary line is less distinct and discontinuous, but here the differences 
cited in the original description cease. The underparts of the Yunnan 
bird are said to be " paler and of a more yellowish tinge." The under 
parts of the series that Rock has sent are more buffy on the breast and 
flanks but could not be spoken of as paler. The bill of the Yunnan bird 
is said in the original description to be smaller but the bills of the series 
at hand are noticeably larger and heavier. 

It may be that the difference in coloration of the under parts is a 
seasonal one since the type of Alcippornis n. yunnanensis Harington 
was taken in April and the series at hand was taken in August and 
September. Unless, however, the birds at hand are freshly molted 
I cannot see that this would account for the more bulTy tone to the 
color. It would seem likely that these birds molt in July or x\ugust 
since La Touche has taken Alcippornis n. hueti David breeding in Fukien 
in late June ^ but there is not a single molting bird in this series. 

'Birds of East China, p. 75. 



144 bulletin: museltm of comparative zoology 

The Yunnan bird is undoubtedly different from its close relati\e 
Alcippornis ti.. fratercula Rippon, the range of which lies to the west 
and southwest, in being more buffy on the breast and flanks and having 
the top of the head more gray and less bro-^Tiish, and from Alcippornis n. 
schacffcri La Touche in having the underparts (breast and flanks) 
much more bufty, schneffcri being lighter below than fratercula. 

Rock has sent a single male from Su-wa-tong, two females and three 
unsexed specimens from Champutong (Chamutang) and two males, 
four females and five unsexed specimens from the Mountains of Tse- 
chung (August). 



Tribura thoracica thoracica (Blyth) 
Dumeticola thoracica Blyth, Journ. Asiat. Soc. Bengal, 14, 1845, p. 584 (Nepal). 

Five males, a single female and two unsexed specimens were taken 
at Su-wa-tong, Tibet, (July), and a single male from the mountains of 
Tung-la (August). 



Neornis flavolivaceus intricatus (Hartert) 

Horeites flavolivaceus intricatus Hartert, Vog. Pal. Faun., 1, 1909, p. 533 (Tai- 
pai-shan, Shensi). 

Four males and four unsexed specimens were taken at Su-wa-tong 
(July). 



Horeites brunnifrons umbraticus Stuart Baker 
Bull. B. O. C, 44, 1924, p. 63 (SchweU-Salween Divide, Yunnan). 
A single male specimen comes from the mountains of Tung-la (To-la) 



(August). 



Phylloscopus .^JiMANDii (Milne-Edwards) 



Abrornis armandii Milne-Edwards, Nouv. Arch. Mas. Paris, Bull., 1, 1865, 
p. 22 (N. China). 

Their large size and the brownish cast to the feathers of the head 
and back distinguish these birds from all others of the group. 

A pair and two unsexed specimens come from Tao-mung-chung 
(April or May). 



GREENWAY: birds from northwest YUNNAN 145 



Phylloscopus subaffinis (Grant) 

Oreopneuste subaffinis Grant, Bull. B. 0. C, 10, 1900, p. 37 (Pu-an-ting, S. W. 
Kweichu). 

This series is as dark or darker below, and lighter above than Phyllo- 
scopus annandii (Milne-Edwards) without the brownish cast to the 
feathers of the back and head. It is distinctly smaller. 

Rock has sent a pair, a female and two unsexed specimens from the 
mountains west of Wei-hsi (June) and two males and two unsexed 
specimens from Tao-mung-chung. 



Phylloscopus fuscatus robustus Stresemann 
Abh. Ber. Mus. Tierk. Volkerk. Dresden, 16, 1924, p. 16 (Sungpan, Sechuan). 

Three specimens have been sent, two unsexed specimens from Tao- 
mung-chung (April or May), and a female from the mountains west 
of Wei-hsi (June). 



Phylloscopus maculipennis debilis (Thayer and Bangs) 

Regidoides rnacidipennis debilis Thayer and Bangs, Mem. Mus. Comp. ZooL, 
40, 1912, p. 180 (Kiating, Sechuan). 

Single male specimens come from Tao-mung-chung (April or May), 
Shwe-men-kan (January) and a female from Mt. Satseto (February). 



Phylloscopus reguloides davisoni (Gates) 
Acanthopneuste davisoni Gates, Faun. Brit. Ind., 1, 1889, p. 420 (Tenasserim). 

I have followed Stuart Baker ^ in retaining the name davisoni in the 
belief that to change the name to flnvolicaceus (Hume) would confuse 
matters unnecessarily. It is a question as to whether fiavolivaceus 
(Hume) is really a SNTionym of Phylloscopus regidoides regidoides (Blyth) 
as Baker believes.^ 

Two unsexed specimens come from Su-wa-tong, Tibet, (July). 

1 Faun. Brit. Ind., 8, 1930. p. 643. 
' Faun. Brit. Ind., 7, p. 189. 



146 bulletin: museum of comparative zoology 

Phylloscopus reguloides claudiae (La Touche) 

Acanthopneuste trochiloides claudiae La Touche, Bull. B. O. C, 43, 1922, p. 22 
(Mengtz, Yunnan). 

Rock has sent nine males, a female and tliree unsexed specimens from 
Tso-mung-chung (April or May); four males and a female from Chou- 
yu-gko (April); a pair and two unsexed specimens from the mountains 
of Tung-la (To-la) (August); a male from Su-wa-tong, Tibet, (July), 
ard an imsexed specimen from the Tse-chung mountains (iVugust or 
September). 

Phylloscopus proregulus forresti Rothschild 
Novit. Zool., 28, 1921, p. 45 (Likiang Range, Yunnan). 

Rock has sent four males, two females and an unsexed specimen 
frcm Tao-mung-chung (April or May), a pair and one unsexed specimen 
frcm Chou-yu-gko (April), and a female from the mountains of Tung-la 

(To -la) (August). 

Phylloscopus nitidus saturatus (Stuart Baker) 

Acanthopneuste nitidus saturatus Stuart Baker, Bull. B. O. C, 44, 1924, p. 62 
(Daban, South Annam). 

A male comes from Chou-yu-gko, a pair from Champutong (Chamu- 
tang) (July), and a male and two unsexed specimens from Su-wa-tong 

(July). 

Phylloscopus affinis (Tickell) 
Motacilla affinis Tickell, Journ. Asiat. Soc. Bengal, 2, 1833, p. 576 (Borabhum). 

These birds are much greener below than PhyUoscopus suhaffinis 
(Grant); they are smaller than PhyUoscopus armandii (Milne-Edwards) 
ard the under wing coverts are whiter than either of them. 

Cne male and one unsexed specimen come from Tao-mung-chung 
(April or May). 

Phylloscopus magnirostris Blyth 
Journ. Asiat. Soc. Bengal, 12, 1843, p. 966 (Calcutta). 

These specimens are darker on the head and back than Phylloscopus 
I. horcalis (Bias.) but they appear to be but little more green below. 

A male and an unsexed specimen were taken at Tao-mung-chung and 
two males and a female at Chou-yu-gko (April). 



GREENWAY: birds from northwest YUNNAN 147 

Phylloscopus pulcher vegetus (Bangs) 

Reguloides pulcher vegetus Bangs, Proc. Biol. Soc. Wash., 26, 1913, p. 95 (Chia- 
kim, West Sechuan). 

A male was taken in the mountains of Tung-la (To-la) (August), 
a pair from Chou-yu-gko (April), and two females and two unsexed 
specimens from Su-wa-tong, Tibet, (July). 

There is insufficient Indian material at hand for comparison. This 
series is \'ery slightly darker and greener below than the type of vegetus. 

Abroscopus schisticeps ripponi (Sharpe) 

Cryptolopha ripponi Sharpe, Bull. B. O. C, 13, 1902, p. 11 (Gyi-dzin-shan, 
Yunnan). 

A pair was taken at ISIt. Satseto in December, four unsexed speci- 
mens come from the mountains west of Wei-hsi (June) and a single 
unsexed bird from Tao-mung-chung, taken in April or May. 

Seicurus burkii distinctus (La Touche) 

Cryptolopha burkii distincta La Touche, Bull. B. O. C, 43, 1922, p. 41 (Mongtz, 
Yunnan). 

Two males, a female and an unsexed specimen come from the moun- 
tains west of Wei-hsi (June), two unsexed specimens from west of 
Wei-hsi (June), a pair from Chou-yu-gko, a male from Tao-mung-chung 
(x\pril or May), an unsexed specimen from the mountains of Tse-chung 
(August or September), and an unsexed specimen from Su-wa-tong, 
Tibet, (July). 

Outram Bangs has published a paper on the burkii form-circle ^ in 
which he has gone into the characters of this group very thoroughly. 
The distinctus group is smaller than Seicurus b. valentini (Hartert). 

Seicurus burkii valentini (Hartert) 

Cryptolopha burkii valentini Hartert, Vog. Pal. Faun., 1, 1907, p. 497 (Tai-pai- 
shan). 

A male from Tao-mung-chung (April or May) appears to be. in- 
termediate between valentini and distinctus. Although it is as large 
as valentini, it has the crown stripes blacker than typical specimens of 
that form, approaching more closely to distinctus. 

Rock has sent two males from Tao-mung-chung and five males from 
Chou-yu-gko (April). 

iProc. N. E. Zool. Soc, 11, 1929, pp. 1-5. 



148 bulletin: museum of comparative zoology 

HoMOCHLAMYS MAJOR (Hors. and Moore) 
Cat. Birds Mus. E. Ind. Co., 1, 1854, p. 323 (Nepal). 

This species has been sent only from Su-wa-tong, Tibet, (July). 
The series consists of only three specimens, two males and one un- 
sexed example. 

CULICICAPA CEYLONENSIS ANTIOXCENTIOR Obcrholscr 
Smithsonian Misc. Coll., 74, July 1923, p. 69 (Tenasserim). 
CuUcicapa c. orienialis Stuart Baker, Bull. B. O. C, 44, Nov. 1923, p. 12. 

Five males and four unsexed specimens were taken at Tao-mung- 
chung in April or May, a single unsexed specimen at Wei-hsi in June, 
a female and an unsexed specimen at Chou-yu-gko in May, an unsexed 
specimen at the mountains of Tse-chung in August or September. 

Chelidorhynx hypoxanthum (Blyth) 

Rhipidura hypoxantha Blyth, Journ. Asiat. Soc. Bengal, 12, 1843, p. 935 
(Darjiling). 

Rock took a single female at Chou-yu-gko in May, seven males, two 
females and four unsexed immature specimens at Su-wa-tong, Tibet, 
in July. He also secured a single unsexed specimen at the mountains 
of Tung-la (To-la) in August. 

EuMYiAS thalassina th.vlassina (Swainson) 

Muscicapa thalassina Swainson, Nat. Lib., 17 (Flycatchers), 1838, p. 252 
(India). 

Four males were taken at Tao-mung-chung (April or May), three 
females and an immature specimen west of Wei-hsi in June and an 
unsexed specimen in the mountains of Tse-chung in August or Sep- 
tember. 

SiPHiA strophiata strophiata Hodgson 
Indian Rev., 1, 1837, p. 651 (Nepal). 

Two males come from Tao-mung-chung (April or May), four 
males and two females from Chou-yu-gko (May), a male and two 
females from Su-wa-tong, Tibet, (July) and a male, three females and 
seven immature specimens in the spotted plumage were taken at the 
mountains of Tung-la (To-la) in August. 



GREENWAY: birds from northwest YUNNAN 149 

SiPHIA PARVA ALBICELLA (Pallas) 
Muscicapa albicella Pallas, Zoog. Russo-Asiat., 1, 1827, p. 462 (Dauria). 

A single specimen, a male, comes from Tao-mung-cliung (April or 
May). 

MUSCICAPULA TRICOLOR TRICOLOR (HodgSOn) 

Digenea tricolor Hodgson, Proc. Zool. Soc. London, 1845, p. 26 (Nepal). 

Two males and three females were taken at Chou-\ u-gko in April 
and a female and two immature specimens at the mountains of 
Tung-la (To-la) in August. 

MUSCICAPULA SAPHIRA Blvth 

Journ. Asiat. Soc. Bengal, 12, 184.3, p. 939 (Sikkim). 

A single male specimen was taken west of Wei-hsi in June. Ander- 
son recorded one male from Ponsee and Forrest sent two males to 
Rothschild. I cannot find that any more than these three birds have 
ever been taken in Yunnan. 

Mltscicapula hyperythra (Blyth) 

Mtisdcapa hyperythra Blyth, Journ. Asiat. Soc. Bengal, 11, 1842, p. 885 
(India). 

Three males were taken at Tao-mung-chung in April or May and a 
single female at Chou-yu-gko in April. 

MusciCAPULA HODGSONii (Verreaux) 

Siphia hodgsonii Verreaux, Nouv. Arch. Mas. Paris, Bull., 6, 1871, p. 34 
(Moupin). 

Of a fine series of thirty-nine birds there are one male, two females 
and four immature birds from the mountains of Tung-la (To-la), 
taken in August, eight males and fifteen females from Chou-yu-gko 
(April), seven males and two females from Tao-mung-chung (April 
or May). 

MusciCAPULA viviDA OATESi (Salvadori) 

Niltava oatesi Salvadori, Ann. Mus. Civ. Genova, 5, 1887, p. 514 (Pegu). 

Rock has sent a good series of three males from Tao-mung-chung 
(April or May), two males and five females from Chou-yu-gko (April), 



150 bulletin: museum of compakative zoology 

and a female and five immature specimens, three of which appear to 
be males, from the mountains of Tung-la (To-la) in August. 

Muscicapula superciliaris aestigma (Gray) 
Muscicapa aestigma Gray, Cat. Mamm. Birds Nepal, 1846, pp. 90, 155 (Nepal). 
Four males come from Tao-mung-chung (April or May). 

Hemichelidon sibirica rothschildi Baker 

Bull. B. O. C, 43, 1923, p. 156 (Yunnan). 

Rock obtained two males from Chou-yu-gko in May, a male, three 
females, an unsexed specimen and two immature birds in the spotty 
plumage from the mountains of Tung-la (To-!a) in August. 

Hemichelidon ferruginea Hodgson 

Proc. Zool. Soc. London, 1845, p. 32 (Nepal). 

Three males, a female and one unsexed specimen have been sent 
from Chou-yu-gko (April). 

NiLTAVA SUNDARA DENOTATA Bangs and Phillips 

Bull. Mus. Comp. Zool., 58, 1914, p. 280 (Mongtz, Yunnan). 

Two males come from west of Wei-hsi (June) and Tao-mung-chung 
(April or May); a pair and an immature male and two females were 
taken at the mountains of Tung-la in August. 

NiLTAVA macgrigoriae (Burtou) 

Phoenicura macgrigoriae Burton, Proc. Zool. Soc. London, 1835, p. 152 (Hima- 
layas). 

A single specimen was taken at Champutong (Chamutang) in July. 

Rhipidura albicollis albicollis (Vieillot) 

Platyrhynchus albicollis Vieillot, Nouv. Diet. d'Hist. Nat., 27, 1818, p. 13 
(Bengal). 

Rock has sent three males, two from the mountains of Tse-chung 
(August or September) and one from Mt. Satseto (December). 



GREENWAY: birds from northwest YUNNAN 151 

Pericrocotus brevirostris styani Baker 
BuU. B. 0. C, 40, 1920, p. 117 (Sechuan). 

Six males and five females come from Tao-mung-chimg (April or 
IVIay) and a male and two females from Chou-yu-gko (April) and a 
single female from the mountains of Tung-la (To-la) (August). 

In his description Stuart Baker says that the males are not dis- 
tinguishable from affinis. This series, however, is much lighter than 
spring birds from Loukouchai and Militi, south Yunnan, being strik- 
ingly lighter below. They resemble brevirostris. The females, how- 
ever, are greener on the back and have much less yellow on the fore- 
head than that form and are otherwise quite like the description of 
sti/ani. 

Lalage melaschista avensis Blyth 

Cat! Birds Asiat. Mus., 1854, p. 327 (Arakan). (Described in Journ. Asiat. Soc. 
Bengal, 15, 1846, p. 307.) 

Two males come from Chou-yu-gko (April). 



Microscelis leucocephalus leucocephalus (Gmelin) 

Turdus leucocephalus Gmelin, Syst. Nat., 1, 1789, p. 826 (China). 

In a series of ten specimens, three males from Tao-mung-chung 
(April or May) one female west of Wei-hsi (June), one male from 
Champutong (July), and one male, two females and two immature 
specimens from the mountains of Tse-chung (August or September), 
there is not one referable to Microscelis I. concolor Blyth. They are slight- 
ly larger, the males measure : wing 121 to 124 mm., as against 115 to 
118 mm. for the male specimens of concolor in the Museum of Com- 
parative Zoology. Furthermore none of them have the blue gray 
sheen to the feathers of the breast that is characteristic of concolor. 
There is a perfect intermediate specimen between concolor and leu- 
cocephalus in the series of the Museum of Comparative Zoology, 
which has the blue gray sheen on the breast as well as a pure white 
head and neck. Rock took leucocephalus in the Likiang Range 
in 1925.1 

1 Proc. U. S. Nat. Mus., 70, 1926, p. 21. 



152 bulletin: museum of comparative zoology 



IXOS MCCLELLANDI SIMILIS (Rothschild) 

lole mcclellandi similis Rothschild, Novit. Zool., 28, 1921, p. 51 (Schweli-Sal- 
ween Divide). 

Comparison of specimens of the same sex of Lros m. holti (Swinh.) 
and Lros m. similis (Roths.) taken in the same months in Fukien 
and Mongtz discloses that the south Yunnan bird has slightly yellower 
under tail coverts. Three specimens, a pair and a bird of the year 
just out of the nest, taken at Champutong in July by Rock, are much 
more clearly referable to similis, being lighter below and having the 
tail coverts much lighter yellow. 

The bills of these three birds are larger than those of either holti 
or similis. This difference is noticeable also in the nestling when 
compared with a nestling in the collections of the Museum of Com- 
parative Zoology, which was taken in August at Mongtz. 



Spizixos canifrons canifrons Blyth 
Journ. Asiat. Soc. Bengal, 14, 1845, p. 571 (Cherra Punji). 

There are no specimens of the Indian or Burmese bird at hand 
for comparison and therefore I cannot say whether this bird has a 
gray or a brown throat. In a long series from south Yunnan and 
Sechuan in the Museum of Comparative Zoology the throats of the 
mature birds are gray, or perhaps brownish gray by a slight stretch 
of the imagination. None are dark brownish gray as described by 
Stuart Baker and so figured.^ 

Since Rothschild ^ considers that the character of the greenness or 
yellowness of the underparts is due to age or individual difference 
and my own comparison of the Chinese specimens in the Museum 
of Comparative Zoology has led me to the same conclusion, I follow 
him in considering the Yunnan bird identical with the Indian and 
Burmese. 

Three birds, two males and a female come from the Likiang Moun- 
tains and were taken in February. A fourth is apparently an immature 
female and was taken in the same place in December. This specimen 
is much darker below than the others. 

"Faun. Brit. Ind., 1, p. 400. 
2 Novit. Zoijl., 28, 1921, p. 50. 



GREENWAY: birds from northwest YUNNAN 153 

Pycnonotus aurigaster xanthorhous Anderson 

Proc. Asiat. Soc. Bengal, 1869, p. 265 (Kakhyen Hills). 

A single male specimen was taken at Tao-mung-chung in April 
or May. 

Lantus tephronotus (Vigors) 
CoUurio tephronotus Vigors, Proc. Zool. Soc. London, 183 1, p. 43 (Himalayas). 

Five males and three females were taken at Tao-mung-chung in 
April or May and three males at Chou-yu-gko in May. 



CONOSTOMA AEMODIUM AEMODIUM HodgSOn 

Journ. Asiat. Soc. Bengal, 10, 1841, p. 857 (Nepal). 

There is a good deal of individual variation in the measurement 
of the wings of this series, the males measuring 121-136 and the 
females 118-134. There is, of course, a good deal of doubt about 
the sexing done by the native collectors that Rock employs, but the 
females seem to be slightly smaller than the males. 

Three males and two females were taken at Tao-mung-chung (April 
or May); three males and a single female at Chou-yu-gko (April); 
three males and one unsexed specimen at Mt. Satseto (December); 
and two pairs at Shwe-men-kan (February). 



Paradoxornis guttaticollis David 
Nouv. Arch. Mus. Paris, Bull., 7, 1871, p. 14 (W. Sechuan). 

Only two females, one from Tao-mung-chung and the other from 
Chou-yu-gko, are in this series. 

SuTHORA unicolor CANASTER Thayer and Bangs 

Mem. Mus. Comp. Zool., 40, 1912, p. 117 (Washan, Sechuan). 

Four males, six females and two unsexed specimens come from 
Mt. Gyi-na-loko (October or November), a single female from Cham- 
putong (Chamutang) (April), two females and an unsexed specimen 
from Shwe-men-kan (January) and a pair from Yung-ning (February). 



154 bulletin: museum of comparative zoology 

SUTHORA FULVIFRONS CYANOPHRYS David 

Joum. Trois Voy. Chine, 1, 1875, p. 345 (Shensi meridional). 

Four males, three females and four unsexed specimens come from 
Tao-mung-chung, two males and three unsexed specimens from 
Chou-yu-gko, and six males, five females and five unsexed specimens 
from Mt. Satseto. 

SuTHORA webbiaxa ricketti (Rothschild) 

Paradoxornis webbiana ricketti Rothschild, Bull. B. O. C, 43, 1922, p. 11 
(Yangtze Valley, Yunnan). 

Three males and a single female come from Mt. Satseto (December). 

SuTHORA WEBBIANA BRUNNEA Ajlderson 
Proc. Zool. Soc. London, 1871, p. 211 (Western Yunnan). 

T have followed Stuart Baker ^ in recording this bird as a subspecies 
of webhiana. Although Rothschild records - it as a distinct species, 
there seems to be no evidence that the two birds have been found 
breeding in the same locality. 

Five adult specimens in the present series are lighter below than a 
single specimen of brunnea from Burma that I have for comparison. 
They are, however, in worn plumage while the single specimen in the 
collection of the Museum of Comparative Zoology is old and probably 
has darkened with time. 

Three males, a female and three immature specimens come from 
west of Wei-hsi and east of the Mekong River (June) and a single 
female comes from Champutong (Chamutang) (July). 

The violet color of the throat in the immature specimens is some- 
what lighter. 

Regulus regulus yunnanensis Rippon 
Bull. B. O. C, 19, 1906, p. 19 (W. Yunnan). 

A male, three females and three unsexed specimens come from Mt. 
Satseto (February). 

1 Faun. Brit. Ind., 7, 1930. p. 21. 
J Novit. Zool., 33, 1926, p. 310. 



GREENWAY: birds from northwest YUNNAN 155 

Aegithaliscus bonvaloti bonvaloti (Oustalet) 

Acredvla bonvaloti Oust., Ann. Sci. Nat. Zool., 12, 1891, p. 286 (Ta-tsien-lu). 

Seven males and three females come from Tao-mung-ehung (April 
or May), two males and one female from Chou-yu-gko (April), four 
males and one female from Mt. Satseto (February), and single males 
from west of Wei-hsi (June), mountains of Tung-la (To-la) (August) 
and Mt. Gyi-na-loko (October or November). 

Aegithaliscus concinnus talifuensis Rippon 
Bull. B. O. C, 14, 1903, p. 18 (Gyi-dzin-shan, N. Shan States). 

Two males were taken at Tao-mung-chung in April or May, and a 
male, two females and two immature specimens from west of Wei-hsi 
in June. 

Stuart Baker remarks ^ that this form is very doubtfully distinct 
from Aegithaliscus c. concinnus (Gould) from China and Yunnan. I 
find that in comparison with ten specimens taken in the spring in 
eastern China (Fukien and Sechuan), the Yunnan birds have the 
buffy brown of the pectoral band and the head a little darker. Birds 
from Mongtz in southern Yunnan seem to be identical with those from 
N, W. Yunnan. 

Sylviparus modestus saturatior Rippon 

Bull. B. O. C, 16, 1900, p. 87 (Mt. Victoria, Chin Hills). 

Three males and two females come from Tao-mung-chung (April or 
May), a male from Su-wa-tong, Tibet, taken in July and an unsexed 
specimen from the mountains of Tung-la (To-la) (August), and two 
males and a female from Mt. Satseto (December). 

Parus dichrous wellsi Baker 
Bull. B. O. C, 38, 1917, p. 8 (Yangtze Big Bend). 

Six males and a female were taken at Chou-yu-gko in April; a single 
male at Tao-mung-chung in April or May; two unsexed specimens at 
Su-wa-tong, Tibet, in July; two males and an unsexed specimen at the 
mountains of Tung-la (To-la) in August ; a male at Mt. Gyi-na-loko in 
October or November, an unsexed specimen at Mt. Satseto in Decem- 

> Faun. Brit. Ind., 1, 1922, p. 95. 



156 bulletin: museum of comparative zoology 

ber; eight males, two females and four unsexed specimens at Shwe- 
men-kan in January and three males and a female at Mt. Satseto in 
February. 

Parus rufonuchalis beavani (Jerdon) 

Lophophanes beavani Jerdon, Birds of India, 2, 1863, p. 275, ex. Blyth MS. (Mt. 
Teringloo, Sikkim). 

Two males and a female were taken at Chou-yu-gko in April; a 
single male at Su-wa-tong, Tibet, in July; a male and four females at 
Mt. Gyi-na-loko in October or November; two males, four females and 
two unsexed specimens from Shwe-men-kan in January and a single 
male from the east slopes of the Likiang Snow Range in February. 

Parus ater aemodius Blyth 
Journ. Asiat. Soc. Bengal, 13, 1844, p. 943 (Nepal). 

A single male comes from Tao-mung-chung (April or May), a male, 
three females and an unsexed specimen from Chou-yu-gko (April); 
an immature specimen comes from west of Wei-hsi in June. A single 
male from Mt. Satseto, taken in February has a wing measurement of 
64 mm. It differs in no other way from the spring specimens and I 
have concluded that it is simply an abnormally large specimen. 

Parus monticolus yunnanensis La Touche 
Bull. B. O. C, 42, 1921, p. 51 (S. E. Yunnan). 

Four males and a female were taken at Tao-mung-chung in April or 
May; three males and two females at Chou-yu-gko in April; a single 
unsexed specimen at Su-wa-tong, Tibet, in July; and a male and three 
unsexed specimens in the mountains of Tung-la (To-la) in August. 

These specimens have the underparts very slightly greener than six 
specimens of the Indian bird in the Museum of Comparative Zoology. 
I can find no other difference. 

Parus palustris dejeani Oustalet 
Bull. Mus. Hist. Nat. Paris, 3, 1897, p. 209 (Ta-tsien-lu). 

Two males come from Tao-mung-chung, a single female from Chou- 
yu-gko (April) and a single male from the Likiang Snow Range 
(February). 



GREEN way: birds FROM NORTHWEST YUNNAN 157 

Parus major altarum La Touche 
Bull. B. O. C, 42, 1922, p. 53 (Yunnan). 

Two specimens, one a female (sexing?) from the Likiang Snow Range 
(February) and one unsexed specimen from west of Wei-hsi (June) 
have been sent. 

Cephalopyrus flammiceps olivaceus Rothschild 
Novit. Zool., 30, 1923, p. 263 (Vicinity of Tengueh). 

Three males were taken at Tao-mung-chung in April or May. 

SiTTA yunnanensis Grant 
Bull. B. O. C, 10, 1900, p. 37 (Wei- Yuan). 

Two unsexed examples come from west of Wei-hsi (June), a male 
from the Likiang Range (February) and a male from Mt. Gyi-na-loko 
(October or November). 

SiTTA europaea nebulosa La Touche 
Bull. B. O. C, 42, 1921, p. 30 (S. E. Yunnan). 

Rock has sent a fine series. Five males and a female come from Tao- 
mung-chung (April or May); three males from Chou-yu-gko (April), 
two males, a female and two unsexed specimens from the mountains 
of Tung-la (To-la) (August); a single male from the mountains of Tse- 
chung (August or September); three unsexed specimens from Su-wa- 
tong, Tibet, (July), a male, a female and an unsexed specimen from Mt. 
Gyi-na-loko (October or November); three males and a female from 
Mt. Satseto in February and a pair from the east slopes of Mt. Satseto 
in ]March. 

The type of nebulosa which is in the ^Museum of Comparative 
Zoology was taken in southern Yunnan (Milati) in February and is 
very gray below as are all the winter birds in the collections of the 
Museum of Comparative Zoology and the specimens in the series that 
Rock has just sent. Spring and summer birds are much lighter and 
browner, less gray. 

TiCHODROMA muraria (Linnacus) 
Certhia muraria Linn., Syst. Nat. edit. 12, 1, 1766, p. 184 (South Europe). 
A single male comes from Mt. Gyi-na-loko (October or November). 



158 bulletin: museum of comparative zoology 

Certhia himalayana yttnnanensis Sharpe 

Bull. B. O. C, 13, 1902, p. 11 (Yunnan). 

Two males, two females and an unsexed specimen come from Tao- 
mung-chung (April or May), a male from the mountains of Tung-la 
(To-la) (July), a single female from Mt. Gyi-na-loko (December) 
and a pair and an unsexed specimen from Mt. Satseto (February). 

Certhia f.^jmiliaris khamensis Bianchi 
Bianchi in Sharpe's Hand List of B., 4, 1893, pp. 355, 360 (Kham). 

A male comes from the mountains of Tung-la (To-la) (August), 
an unsexed immature bird from Su-wa-tong, Tibet, (July), two males 
and an unsexed specimen from Mt. Gyi-na-loko (October or Novem- 
ber) and a male from the eastern slopes of the Likiang Snow Range 
(February). 

ZosTEROPS simplex SIMPLEX Swinhoe 

Proc. Zool. Soc. London, 1862, p. 317 (S. E. China). 

A male, seven females and five unsexed specimens were taken at 
the mountains west of Wei-hsi in June and a single unsexed example 
at Su-wa-tong, Tibet, in July. 

Pachyglossa melanoxantha Hodgson 
Journ. Asiat Soc. Bengal, 12, 1843, p. 1010 (Nepal). 

A male and two females were taken at Chou-yu-gko in April. 

Aethopygia ignicauda exultans Baker 
Bull. B. O. C, 46, 1925, p. 13 (Schweli-Salween Divide, w. central Yunnan). 

A series of twelve males, an unsexed specimen (young male) and 
six females were shot at Su-wa-tong, Tibet, in July. 

Aethopygia dabryi (Verreaux) 

Nedarinia dabryi Verreaux, Rev. et Mag. Zool., 1867, p. 173 (Ta-tsien-lu, 
Sechuan). 

Eight males and five females were taken at Tao-mung-chung in 
April or May. Five pairs were taken at Chou-yu-gko in April, a pair 



GREENWAY: birds from northwest YUNNAN 159 

at Champutong (Chamutang) in July and a single male at Wei-hsi 
in June; a young bird just out of the nest was taken at Su-wa-tong, 
Tibet, in July. 

MOTACILLA ALBOIDES HodgSOn 

Asiat. Res., 19, 1836, p. 191 (Nepal). 

Rock has sent three males and two females from Tao-mung-chung 
(April or IVIay). I have followed Stuart-Baker in considering all 
forms of the black backed group as distinct species. 

MOTACILLA LEUCOPSIS Gould 

Proc. Zool. Soc, London, 1837, p. 78 (India). 

A single specimen comes from Tao-mung-chung (April or May). 
It is a male. 

MoTACiLLA ciNEREA CASPiCA (Gmelin) 

Farm caspicus Gmelin, Reise Russ. Reichs., 3, 1774, p. 104, pi. 20, fig. 2 (Cas- 
pian Sea). 

A single male specimen was sent from Tao-mung-chung (April or 
May). It is just beginning to molt out of winter plumage. 

Anthus hodgsoni yunnanensis Uchida and Kuroda 
Annot. Zool. Japan, 9, 1916, p. 134 (Yunnan). 

Four males and three females were taken at Chou-yu-gko in May 
and two males at Mt. Gyi-na-loko in October or November. The 
autumn birds are very much darker. The back is green not grayish 
green and the yellow of the throat is darker and more buffy. 

Anthus roseatus Blyth 
Journ. Asiat. Soc. Bengal, 16, 1847, p. 437 (Nepal). 

Rock took two pairs at Tao-mung-chung in April or May. 

Alauda gulgula weigoldi Hartert 

Abh. u. Ber. d. Zool. Anthr. Ethn. Mus. zu Dresden, 15, 1922, no. 3, p. 20 
(Hankow). 

A male, a female and an unsexed specimen were taken at Tao- 
mung-chung in April or May. Their wings measure 101, 101 and 106 



160 bulletin: museum of comparative zoology 

. . . the extreme of size given for weigoldi. I can find very little 
diflFerence in size between the short-winged larks of south and north 
China. 

I prefer to follow Stuart Baker in retaining gulgula for the short- 
winged birds. 

Emberiza pusilla Pallas 
Reise Russ. Reichs., 3, 1776, p. 697 (Daurian Alps). 

Six males, a female and one unsexed specimen come from Tao- 
mung-chung (April or May). 

Emberiza fucata arcuata Sharpe 

Cat. Birds Brit. Mus., 12, 1888, p. 494 (Himalayas). 

A pair was taken in the mountains west of Wei-hsi in June. They 
are in worn breeding plumage. They are as dark as a coty pe of Emberiza f. 
kuatunensis La Touche (taken in April), but the chestnut borders of 
the feathers of the median coverts and the lesser coverts are worn 
and faded so that the appearance of the bird is somewhat changed. 

Emberiza spodocephala melanops Blyth 
Journ. Asiat. Soc. Bengal, 14, 1845, p. 554 (Tippera). 

Four males and a female (sexing?) were taken west of Wei-hsi in 
June. 

The breeding range that Sushkin has given ^ for this bird should be 
extended to include Yunnan. In spite of the fact that there is no 
direct evidence that these birds were breeding in western Yunnan, 
it is very probable that they were, since they were taken there in June. 
For this reason it may not be that all birds breeding west of Sechuan 
are Emberiza s. spodocephala (Pallas). 

Emberiza elegans elegantula Swinhoe 

Proc. Zool. Soc. London, 1870, p. 134 (Kweichow, Hupeh). 

Eleven males and three females (young males?) were taken at 
Tao-mung-chung in April or May, two pairs and four immature 
specimens (unsexed) from west of Wei-hsi in June; a single male comes 

' Proc. Soc. Nat. Hist., Boston, 38, 1925, p. 28. 



GREENWAY: birds from northwest YUNNAN 161 

from the Likiang Snow Range in March and a single immature speci- 
men from Champutong in July. 

A single nestling specimen of what appears to be this form was 
taken at Su-wa-tong, Tibet, in August. 

Emberiza aureola Pallas 
Reise Russ. Reichs., 2, 1773, p. 711 (Irtysh). 

Two males and two unsexed specimens (females?) were taken at 
Tao-mung-chung in April or May. 

Emberiza godlewskii yunnanensis Sharpe 
Bull. B. O. C, 13, 1902, p. 12 (TaUfu, Yunnan). 

Three males come from Tao-mung-chung (April or May) ; two males 
from Chou-yu-gko (May) and three males and an unsexed specimen 
from Mt. Satseto (February). 

I have followed Sushkin in his treatment of this group. From a com- 
parison of the specimens in the Museum of Comparative Zoology it 
would appear that he is right in considering the cia group, character- 
ized by a black stripe behind the ear coverts, and the godlewskii 
group with the comparatively broad chestnut stripe, as distinct 
species. 

Passer rutilans intensior Rothschild 
Bull. B. O. C, 43, 1922, p. 11 (Mekong Valley). 

Two pairs were taken west of Wei-hsi in June; two males at Tao- 
mung-chung in April or May, and single females at Champutong in 
July and at Mt. Satseto in December. 

Fringilla montifringilla Linnaeus 
Syst. Nat., 10th Ed., 1, 1759, p. 179 (Sweden). 

Two females were taken at Mt. Satseto in October or November, 
and two females and an unsexed specimen at Tao-mung-chung in 
April or May. 

Procarduelis nipalensis intensicolor Baker 
Bull. B. O. C, 45, 1925, p. 92 (Yunnan). 

Three males come from Mt. Gyi-na-loko (October or November), 



162 bulletin: museum of comparative zoology 

three pairs from Mt. Satseto (December and February), and eleven 
females from Su-wa-tong, Tibet, (July). 

Procarduelis rubescens saturatior Rothschild 
Bull. B. O. C, 43, 1922, p. 12 (Likiang Range, Yunnan). 

Five males and five females come from Shwe-men-kan (January), a 
male from Mt. Satseto (January), a male from Mt. Gyi-na-loko 
(October or November) and a male from Mt. Satseto (January). 

Propyrrhula subhimachala intensior Rothschild 
Bull. B. O. C, 43, 1922, p. 12 (Likiang Range). 

A single specimen, male, was taken at Mt. Gyi-na-loko in Novem- 
ber or December. 

Erythrina thura feminina (Rippon) 
Car-podacus femininus Rippon, Bull. B. O. C, 19, 1906, p. 31 (Western Yunnan). 

Rock has sent a fine series of thirty-seven skins. Three males and 
ten females were taken at Mt. Gyi-na-loko in November and Decem- 
ber; five males and eleven females at Shwe-men-kan in January and 
two males and five females on the east slopes of the Likiang Snow 
Range near Mt. Satseto in January or February and a single female 
from the mountains of Tung-la (To-la) in July. 

Females from the same locality have the rump a very deep reddish 
brown or a rather lighter brown. This seems to me to be an individual 
variation. 

Erythrina verreauxi (Dav. and Oust.) 
Propasser verreauxi Dav. and Oust., Ois. de la Chine, 1877, p. 355 (China). 

A pair was taken at Mt. Gyi-na-loko in October or November; three 
males and two females at Mt. Satseto in December and a male and 
two females in February; single females come from Shwe-men-kan 
(January), Su-wa-tong, Tibet, (July), and the mountains of Tung-la 
(To-la) (August). 

I have followed Berlioz in relegating the name Erythrina ripponi 
(Sharpe) to synonymy. 



GREENWAY: birds from northwest YUNNAN 163 

Erythrina pulcherrima argophrys (Berlioz) 

Carpodacus argophrys Berlioz, Bull. Mus. Paris, ser. 2, 1, 1929, p. 131 
(Ta-tsien-lu and Tseku, Sechuan). 

A pair was taken on the Likiang Range in February; single females 
on the slopes of ]Mt. Gyi-na-loko in October or November, Mt. Satseto 
in ]March and Tao-mung-chung in April or May. 

I have followed Stresemann ^ in the identification of the pulcherrima 
group. He has apparently examined most of the known material in 
Europe. 

Erythrina eos Stresemann 
Orn. Monats., 38, 1930, p. 75 (Sungpan, Sechuan). 

A single specimen was taken in the mountains of Tung-la (To-la) in 
August. 

The wing tip of this specimen measures 12 mm. and the wing meas- 
ures 74 mm., as against 24 mm. and 87 mm. for the specimens that I 
have identified as Erythrina pulcherrima argophrys Berlioz. 



Erythrina edwardsii rubicunda subsp. nov. 

Typc.—M C. Z. No. 159,303, taken at Su-Wa-Tong, Tibet, on the 
upper slopes of the Mt. Kenichunpo or Gomba-La, east slopes of the 
Salween-Irawaddy Divide, 14,000 ft. 

Closest to Erythrina edwardsii saturatus (Blanford) but the head is 
maroon rather than pink with greenish reflections; edges of the 
feathers of the back are reddish brown, without the hint of green; the 
throat is a more purplish red and the breast is maroon only very 
slightly lighter than the head; the belly is pink but without any yel- 
lowish tinge. The effect is such that the bird is noticeably darker 
and redder than the Indian Bird. The female also is much darker 
above, the dark shafts of the feathers being so close together that it 
appears almost to be black. Below, the shafts of the feathers are 
black rather than dark brown and the edges of the feathers are a 
darker brown. 

Description.— Head maroon; a broad light pink stripe running from 
just above and before the eye to the neck; lores dark brown; ear coverts 
maroon like the head; feathers of the back with a broad black shaft 
stripe, the edges reddish brown; tail coverts maroon like the head; tail 

' Orn. Monats., 38, 1930, p. 72. 



f 



164 bulletin: museum of comparative zoology 

feathers dull brown, the outer webs of the outer three narrowly edged 
with reddish brown. Below, brown; throat and breast purplish red 
with black shaft lines to the feathers; belly rose pink; thighs have a 
brownish tinge, giving a bronze effect; primary coverts brown, the 
outer webs narrowly edged with reddish bronze and tipped with pink; 
primaries and secondaries brown, the outer webs narrowly edged with 
reddish bronze. 

Rothschild ^ remarked as follows on a specimen of Erythrina ed- 
wardsii saturat us (BlanL) iroxn Yunnan: "The adult male is much 
darker than any of my Chinese birds, and is even darker than any 
specimen from Nepal and Sikkim that I possess, so it may prove to 
be a third race when more Yunnan material comes to hand." 

Rock has sent a series of four males and eight females from Su-wa- 
tong, Tibet, taken in July and a series of one male and eight females 
taken in the Likiang Snow Range in the vicinity of Mt. Satseto at 
altitudes varying from 9,000 to 14,000 ft. in October, November, De- 
cember, January, February. 



Erythrina vinacea rubidior subsp. nov. 

Type No. 159,258 Museum Comparative Zoology from the moun- 
tains of Tung-la (To-la). 

This bird is like Erythrina v. vinacea (Verreaux) but darker. The 
deeper, almost blackish, red is more noticeable on the upper breast. 
The female is markedly darker than females from Sechuan. 

Rock has sent a pair from the mountains of Tung-la (To-la) taken 
in July, a female from Su-wa-tong, taken in iVugust and a male from 
Mt. Gyi-na-loko, taken in October or November. 

These specimens have been compared with three males and two 
females from Sechuan in the collections of the Museum of Comparative 
Zoology, one male from the La Touche collection without data other 
than the locality "Sechuan," a male from Washan, Sechuan, (May), 
and another from Chetze, Hupeh, (December) ; both females are from 
Washan. 

If the type locality of Erythrina v. vinacea (Verr.) is really "Moun- 
tains of Chinese Tibet" as Hartert and Rothschild have recorded it, 
it would seem very improbable that we should now find a race on 
the border of Tibet and Yunnan. I think, however, that the type 
locality is actually Ho-pa-tchang, "a journey of a day and a half 

1 Novit. Zool., 33, 1926, p. 332. 



GREENWAY: birds from northwest YUNNAN 165 

north of Chengtu, Sechuan," and that David shot his birds in that 
district only.^ I believe that birds from Kansu, Yunnan and eastern 
Tibet are geographical representatives, a western race of vinacea. 



Carpodacus ery'Thrina roseatus (Blyth) 

Propasser roseatus Blyth, (ex Tickell) Journ. Asiat. Soc. Bengal, 11, 1842, p. 
401 (Calcutta). 

Two males and a female come from Tao-mung-chung (April or 
May). 

Pyrrhula erythaca altera Rippon 
Bull. B. O. C, 19, 1906, p. 19 (West Yunnan). 

Rock has sent a fine series of thirty -one skins. A male and six 
females come from Tao-mung-chung (April or May); single females 
were taken at Chou-yu-gko (May) and the mountains of Tung-la 
(To-la) in August; four males and fourteen females come from Mt. 
Satseto (January or February). Two males and three females from 
Mt. Gyi-na-loko (October or November). 

The throat of one of the males taken at Mt. Satseto is suffused with 
red and another has the breast quite green with only a few incon- 
spicuous reddish-orange feathers. None of these specimens have 
the breast vermilion as has a single topotype of Pyrrhula e. taipaisha- 
nensis Roths, in the collections of the Museum of Comparative 
Zoology. There are specimens from Sechuan, apparently immature, 
taken in September in the La Touche collection which have the breast 
suffused with greenish-yellow as the single specimen that Rock has 
sent. I am, therefore, inclined to believe that this coloration is due 
to age or perhaps to the character of the food or to a combination of 
both. 

Pyrrhula nipalensis ricketti La Touche 

BuU. B. O. C, 16, 1905, p. 21 (Mountains of N. W. Fukien). 

A male and two females were taken in the mountains of Tung-la 
(To-la) and a single female in the mountains of Tse-chung, all in 
August. 

> Nouv. Arch. Mus. Paris, Bull., 7, 1871, p. 61; ibid, 8, 1873, pp. 22-26. 



166 bulletin: museum of comparative zoology 

Pyrrhoplectes epauletta (Hodgson) 
Pyrrhula epauletta Hodgs., Asiat. Res., 19, 1836, p. 156 (N. and C. Nepal). 

Twelve males and four females come from Su-wa-tong, Tibet, (July). 
These birds appear to be identical with the specimens from the 
Himalayas in the Museum of Comparative Zoology. 

Spinus thibetanus (Hume) 
Chnjsomitris thibetanus Hume, Ibis, 1872, p. 107 (Sikkim). 

A pair was taken at Mt. Gyi-na-loko in October or November. 

I cannot find that this bird has ever been taken in Yunnan before. 



Hypacanthis ambiguus (Oustalet) 
Chrysomitris ambigua Oustalet, Bull. Mus. Paris, 2, 1896, p. 186 (Yunnan). 

Two females come from Tao-mung-chung (April or May) and four 
males, a single female from Mt. Satseto (February) and an immature 
specimen from Champutong (July). 

The differences in color pattern between this bird and Hypocanthis 
s. spinoides (Vigors) are so great that I prefer to consider them as 
distinct species. 

Perissospiza icteroides affinis (Blyth) 

Hesperiphona affinis Blyth, Journ. Asiat. Soc. Bengal, 24, 1855, p. 179 (Alpine 
Punjab). 

Rock has sent a series of nine birds. A male was taken at Tao- 
mung-chung in April or May and one on the mountains of Tung-la 
in August; four males and two females at Mt. Gyi-na-loko in October 
or November and a single female at Shwe-men-kan in January. 

Perissospiza carnipes carnipes (Hodgson) 
Coccothraustes carnipes Hodgson, Asiat. Res., 19, 1836, p. 151 (Nepal). 

A mature and two immature males and a female were taken at 
Chou-yu-gko in May; a pair at the mountains of Tung-la (To-la) in 
Auiiust and three females on Mt. Satseto in February. 



GREEXWAY: birds from northwest YUNNAN 167 

Mycerobas melanoxanthos (Hodgson) 

Coccothraustes vielanoxanthos Hodgson, Asiat. Res., 19, 1836, p. 150 (Nepal). 

A single specimen, which the collector could not sex, but which 
appears to be a female comes from the mountains of Tung-la (To-la) 
(August). 

Oriolus chinensis tenuirostris Blyth 

Journ. Asiat. Soc. Bengal, 15, 1846, p. 48 (Central India). 

Three males and two females come from west of Wei-hsi (June), 
a single male from Tao-mung-chung (April or May) and a single 
female from the Likiang Snow Range (March). 

Oriolus trailli (Vigors) 
Pastor trailli Vigors, Proc. Zool. Soc. London, 1831, p. 175 (Darjiling). 

A mature, an immature male and a female were taken at Chou- 
yu-gko in April. 

DicRURUS macrocercus cathoecus Swinhoe 

Proc. Zool. Soc. London, 1871, p. 377 (China). 

A single specimen, a male, was shot at Tao-mung-chung in April 
or May. 

DiCRURUS leucophaeus hopwoodi Baker 

Novit. Zool., 25, 1918, p. 294 (Decca). 

Five males and two females were taken at Tao-mung-chung in 
April or May and a single female west of Wei-hsi in June. 

Acridotheres cristatellus cristatellus (Gmelin) 
Gracula cristatella Gmelin, Syst. Nat., 1, 1788, p. 397 (China). 

A single male was taken in the Likiang Snow Range in February. 

Garrulus glandarius sinensis Swinhoe 
Proc. Zool. Soc. London, 1871, p. 381 (South China, westwards to Sechuan). 

Five females (sexing ?) : three were taken at Mt. Gyi-na-loko in 
October or November, one at Mt. Satseto in December and one at 
Shwe-men-kan in Januarv. 



168 bulletin: museum of comparative zoology 

NuciFRAGA caryocatactes macella Thaver and Bangs 
Bull. Mus. Comp. Zool., 52, 1909, p. 140 (Hsien-shan-hsien), 

Three males and a female taken at Chou-yu-gko in April are im- 
mature, another female taken at the same place at the same time is in 
worn plumage and the feathers of the head and back and breast have 
almost lost their white tips and the brown is faded. A female taken 
at Tao-mung-chung in April or May is in fresher plumage and a 
female from Champutong, taken in July is in full plumage and is 
very dark as are two females that were taken at Mt. Gyi-na-loko in 
October or November. 

I cannot find that these birds differ from macella in any important 
characteristic and I believe that Nucifraga c. yunnanensis Ingram ^ 
is probably a synonym. 

> Bull. B. O. C, 25, 1910, p. 86. 



JUL 11 1933 



o-Z^f 



Bulletin of the Museum of Comparative Zoology 

AT HARVARD COLLEGE 
Vol. LXXIV, No. 6 



NEW AND LITTLE KNOWN SPIDERS 
FROM THE UNITED STATES 



By Elizabeth B. Bryant 



With Four Plates 



CAMBRIDGE, MASS., U. S. A. 

PRINTED FOR THE MUSEUM 

June, 1933 



PUBLICATIONS 

OF THE 

MUSEUM OF COMPARATIVE ZOOLOGY 
AT HARVARD COLLEGE 



There have been published of the Bulletin Vols. I to LXV, LX VII- 
LXXIV, of the Memoirs Vols. I to LI. 

The Bulletin and Memoirs are devoted to the publication of 
original work by the Oflficers of the Museum, of investigations carried 
on by students and others in the different Laboratories of Natural 
History, and of work by specialists based upon the Museum Collec- 
tions and Exploration. 

These publications are issued in numbers at irregular intervals. 
Each number of the Bulletin and of the Memoirs may be sold sepa- 
rately. A price Ust of the publications of the Museum will be sent on 
application to the Director of the Museum of Comparative Zoology, 
Cambridge, Massachusetts. 



Bulletin of the Museum of Comparative Zoology 

AT HARVARD COLLEGE 

Vol. LXXIV, No. G 



NEW AND LITTLE KNOWN SPIDERS 
FROM THE UNITED STATES 



By Elizabeth B. Bryant 



With Four Plates 



CAMBRIDGE, MASS., U. S. A. 

PRINTED FOR THE MUSEUM 

June, 1933 



No. f). — New and Little Knoion Spiders from the United States 
By Elizabeth B. Bryant 

Recently while identifying spiders from various collections at the 
ISIuseum of Comparative Zoology several new species and new records 
of distribution were found. In recent years Mr. \V. S. Blatchley has 
sent Professor Banks collections from Dunedin and Royal Palm Park, 
Florida. It is not surprising to find from the latter locality four West 
Indian species, Thcridion rufiiJes Koch, a cosmopolitan tropical species, 
Trachclas bicolor Keyserling, described from Haiti, Corinna (jracilijjes 
(Keyserling) also known only from Haiti, and Pseudos-parianthis 
cidjana Banks described from Havana, Cuba. From Brownsville, 
Texas, were found two Mexican species, M etaphidippiis longipalpus 
Cambridge and Miayrammopes lineatus Cambridge, the first time that 
the latter genus has been found in the United States, and also Myrme- 
coti/pus cuhanus Banks, a Cuban species originally found with ants at 
Soledad, Cuba. 

The following species are new: 

Euryopis emertoni Xysticus laticeps 

Eurijopis ornata Xysticus trimaculatus 

Gonatium crassipalpus Philodromus montanus 

Dipoena lineatipes Philodromus bilineatus 

Trachelas laticeps Philodromus emertoni 

Eho cockerelli 



ULOBORIDAE 

MiAGRAMMOPES LINEATUS Cambridge 

Biol. Centr. Amer., Arachn., 1894, 1, p. 137, pi. 17, fig. 12. 

Among a small lot of spiders collected by C. Schaeffer at Browns- 
ville, Texas, was a small M la gr a mm opes that probably is this species. 
It agrees with a specimen from Costa Rica in size and arrangement of 
eyes. It is the first time that the genus has been recorded from the 
United States. 



172 bulletin: museum of comparative zoology 



THERIDIIDAE 
EuRYOPis emertoni spec. nov. 

Plate 1, fig. 1 

cf 2 mm. long. 

Cephalothorax bright orange red, black about the eyes with a few 
short, stifp hairs in front of the quadrangle of eyes; abdomen, basal 
four-fifth covered with a brown scutum with many long, slender hairs 
from coriaceous spots ; legs bright yellow with many fine hairs ; sternum 
and coxse orange; sternum prolonged in an obtuse lobe between the 
posterior coxae; venter covered with two thickened pieces, one from the 
pedicle to the spiracles and the other continued to the spinnerets, 
nearly the width of the venter; eyes, anterior row recurved, A.M.E. 
slightly largest, separated by diameter and a half and from A.L.E. by 
less than a diameter; posterior row straight, only very little longer than 
anterior row, eyes subequal and equidistant, lateral eyes almost 
touching; quadrangle of median eyes widest in front and wider than 
high; clypeus about three times as high as eye area. 

Palpus small, tibia less than half as long as wide, with simple 
rounded sides and no lobes; palpal organ very simple as figured. 

Holotype (d^) Ga.; Thompson's Mills, H. A. Allard; N. Banks Coll. 

This species is very close to Euryopis spmigera Cambridge, Biol. 
Centr. Amer., 1895, 1, p. 146, pi. 19, fig. 2, but difters as follows: the 
clypeus three not four times the e^'e area, a distinct scutum between 
the epigastric fold and the spinnerets, instead of scattered patches, and 
in the male palpus the narrow tibia without lobes, where in E. spinigera 
it is described and figured as " very short and spreading, produced con- 
siderably on the inner side in an obtuse form." 

It is very probable that the two specimens referred to as E. spinigera 
by Mr. J. H. Emerton, in Psyche, 1924, 31, p. 142, fig. 4 from Chatham, 
Massachusetts, and Charleston, South Carolina, are tliis species. 

Euryopis ornata spec. nov. 
Plate 1, figs. 2, 3 

cf 1.5 mm. long. 

Cephalothorax dull gray brown, black around eyes; abdomen 
brownish gray with six transverse clear bands, those at the apex bent 
almost like chevrons; legs pale yellow without marks; sternum, venter 
and coxie pale yellow; IV coxse separated by more than a diameter; 



BRYANT: NEW AND LITTLE KNOWN SPIDERS 173 

eyes, A.M.JE. largest of the eight, separated by about a diameter; 
lateral eyes almost touching; quadrangle of median eyes widest in 
front; clypeus more than twice as high as quadrangle. 

Palpus very simple; tibia about as long as patella. 

Holotype (d^), Miss.; Meridian, H. E. Weed, N. Banks Coll. 

Euryopis ornata is much smaller than any described American 
species. It differs from E. spinigcra Cambr. reported by Mr. Emerton 
from Chatham, INIassachusetts, and Charleston, South Carolina, by the 
palpus and the lack of a scutum on dorsum and venter, and from E. 
funebris (Hentz) by the much smaller size, unmarked legs and palpus. 

Theridion quadrimaculatum (Banks) 
Plate 1, fig. 4 
Mysmena quadrimaculata Banks, Trans. Amer. Ent. Soc, 1896, 23, p. 66. 

cf 1.5 mm. long. 

Cephalothorax, legs and sternum pale yellow; abdomen yellowish 
gray with four dark blotches as in the female, with many long hairs; 
venter dark in the center; sternum as broad as long; eyes; anterior row 
recurved, A.M.E. separated by two diameters and by one diameter 
from A.L.E.; posterior row, eyes subequal, procurved so that lateral 
eyes touch, P.IM.E. separated by more than a diameter; quadrangle 
about as high as wide; clypeus a little higher than quadrangle of eyes. 

Palpus longer than I femur, patella longer than tibia; palpal organ 
very simple. 

Holotype (9 ) Fla.; Punta Gorda, N. Banks Coll. 

Allotype (d^) Fla.; Dunedin, March, 1927, W. S. Blatchley Coll. 

The type of this species is a female and it was placed in the genus 
Mysmena. In 1925 Mr. Crosby examined the type and said that it did 
not belong to that genus. The male palpus shows that it is a Theridion 
in the sense that Simon uses the genus. 

Theridion rufipes Koch 

This cosmopolitan tropical spider has been found recently at Sebas- 
tian, Florida. 

Gonatium crassipalpus spec, no v. 

Plate 1, figs. 5, 6, 8, 9, 10 

cf 2.5 mm. long; ceph. 1.2 mm.; abd. 1.7 mm. 

Cephalothorax bright brown, black around the eyes; abdomen gray; 



174 bulletin: museum of comparative zoology 

legs orange; head elevated in eye area, sternum brown, widest between 
I coxse and almost as wide as long, extending in a square lobe between 
IV coxse, which are separated by a diameter; mandibles with three equal 
teeth on superior margin of fang groove, I femur with row of stiff 
bristles from coriaceous pits, patella much swollen on upper side, tibia 
bent and dilate near apical end with an irregular mass of dark hairs 
and a fringe of long, colorless hairs, metatarsus with two rows of short, 
stiff bristles which continues on the tarsus; eyes subequal with laterals 
almost touching. 

Palpus ; femur much enlarged, the granulations less numerous than in 
Gonatium rubens, patella longer than tibia, superior apophysis of tibia 
long, slender and curving, almost reaching the tip of the tarsus; below 
are two apophyses, short and black, the lower one quite complicated; 
palpal organ fills the cavity and very complicated. 

9 3.2 mm. long; ceph. 1.5 mm.; abd. 2.2 mm. 

Cephalothorax brown, darker than the male, abdomen bro\\Tiish 
gray; legs brown, about the same color as the cephalothorax with 
many hairs but not in rows as in the male ; sternum brown ; head not 
high as in the male, but slightly elevated and the eyes a little more 
separated. 

Epigynum a large dark cavity, twice as wide as high with the open- 
ings widely separated. 

Holotype (cf) Colo.; Long Lake, 28 August, T. D. A. Cockerell; 
N. Banks Coll. 

Allotype ( 9 ) Colo. ; Long Lake, 28 August. 

Paratype 1 cf Colo.; Boulder, Peaceful Valley, 8,000 ft., August, 
T. D. A. Cockerell; N. Banks Coll. 

This species is less brightly colored than Gonatium rubens (Blackw.), 
the European species found in the east, and differs from it in the male 
palpus and the epigynum in the female. The long, curved apophysis 
on the tibia of the male palpus is longer and more slender and the other 
apophyses are more complicated; the epigynum of Gonatium rubens 
is higher than wide and the cavity is much smaller. 

DiPCENA LINEATIPES spcC. nov. 

Plate 1, fig. 7 

9 1.4 mm. long. 

Cephalothorax yellow, black in eye area and a dark triangle from the 
posterior eyes to the cephalic groove; abdomen round, gray and covered 
with short, stiff, curved hairs from distinct coriaceous granulations. 



BRYANT: NEW AND LITTLE KNOWN SPIDERS 175 

more numerous at base than at apex; two pairs of muscle spots, small 
but distinct; legs yellow, with rows of hairs on tibia and metatarsus 
longer than the diameter of the joint and a broad dark stripe on tibiae 
and metatarsi, more distinct on I and II legs; sternum and coxse 
yellow; venter gray with stiff hairs, smaller than on the dorsum; eyes, 
anterior row straight, A.M.E. separated by more than a diameter, 
about twice as large as A.L.E. and separated from them by a radius; 
posterior row procurved, equidistant, P.M.E. slightly larger than 
P.L.E. and a little smaller than A.M.E. ; lateral eyes touching; clypeus 
about twice as high as quadrangle of eyes. 

Epigynum rather large for the size of the spider, showing two large 
circular cavities partly separated by a septum and a pair of small dark 
dots close together near the epigastric fold. 

Holotype (9) Fla.; Royal Palm Park, 15-24 March, 1930, W. S. 
Blatchley. 

The small size, striped legs and large epigynum are distinct from any 
described species. 

Epeiridae 

The American species of Pachygnatha have been in a confused 
state for many years. In 1845 C. Koch described two species from 
Pennsylvania, P. tristriafa and P. xantfiostoma. These are now in the 
Berlin Museum. In 1882 Keyserling redescribed a male and female 
from Boston, Massachusetts, in the Simon Collection as P. tristriata. 
A year later he decided that it was a new species and called it brevis 
and a species found "at Long Island, Philadelphia, and Columbus, 
(Texas) " in the Marx Collection was called P. tristriata. In the same 
paper he redescribed and figured P. xanthostovm from Philadelphia 
but he evidently did not feel quite sure of his identification for he 
queried it. Later McCook in American Spiders, 1893, 2, p. 269, pi. 26, 
figs. 7, 8 also described P. xanthostovia and in a footnote calls attention 
to the fact that his specimens do not agree in size with that of the orig- 
inal description. He has also added to the confusion by exchanging 
the figures of the mandibles of P. xant/iostoma and P. dorothea McCook. 
The latter, he states, has a tooth on the fang and the artist shows a 
tooth on the upper side of the mandible above the insertion of the fang. 

In 1912, Mr. Banks examined the Koch types at the Berlin Museum 
and found that the species labelled P. xanthostovia is what has been 
recognized as P. brevis. Koch compares his species to the European 
species P. degeeri which also has a short abdomen (but little longer than 



176 bulletin: museum of comparative zoology 

the cephalothorax) and mandibles at least half as long as the cephalo- 
thorax. It also is the most common species in the eastern United States. 

Among the many specimens of P. xanthostoma Koch {hrems Keys.) 
were several males that differ in the arrangement of teeth on the 
superior margin of the mandible and the slender tarsal appendage of 
the palpus. These have been recognized as P. furcillata Keys., a 
species that has been placed as a synonym of P. xanthostoma Koch 
(brevis Keys.). 

P. dorothca McCook is easily recognized by the tooth on the upper 
side of the mandible above the insertion of the fang. The type of the 
genus, P. clcrckii Sund. also has this tooth, but the three European 
species differ from the American by the quadrangle of median eyes, 
which in the European is narrower in front and the P.M.E. always the 
largest of the eight. 

PaCHYGNATHA DOROTHEA McCook 

Plate 4, figs. 37, 41 
American Spiders, 1893, 3, p. 270, pi. 26, figs. 3, 4. 

cf 3.2 mm. long. 

Cephalothorax brown with a broad marginal stripe and a median 
stripe in front of thoracic groove; folium dull brown with irregular 
clear white paired marks, sides and venter pale brown; legs pale yellow; 
eyes, anterior row recurved, A.M.E. carried forward and nearer each 
other than to A.L.E., posterior row slightly procurved, P.M.E. nearer 
each other than to P.L.E., quadrangle slightly higher than wide, 
median eyes subequal, and lateral eyes touching; clypeus not as high 
as quadrangle ; mandibles divergent, about one half as long as cephalo- 
thorax, superior margin of fang groove with a blunt hook or tooth at 
base of fang and three equidistant teeth, inferior margin with four sub- 
equal and equidistant teeth, fang long and sinuate; head relatively low. 

Palpus, femur as long as mandible, tibia a little longer than patella, 
tarsal appendage with a blunt tip, and the usual widened portion one- 
third from the base. 

9 3.5 mm. long. 

The female has the same coloring as in the male, but the clypeus is 
a little higher and about equals height of quadrangle. The mandibles 
are geniculate, not as divergent as in the male and the blunt tooth 
above the base of the fang is missing; the teeth on the inferior margin 
is the same. 

The type of this species was found near Philadelphia. In the museum 



BRYANT: NEW AND LITTLE KNOWN SPIDERS 177 

collection there are specimens from Ithaca, New York; Salineville, 
Ohio; Boulder, Colorado; and INIanitoba, Canada. Because it is a small 
species it has been confused with P. autumnalis. 

Pachygnatha furcillata Keyserling 
Plate 4, figs. 38, 39 
Verb. z. b. Ges. Wien, 1SS3, 33, p. 662, pi. 21, fig. 11. 

d^ 4.8 mm. long; ceph. 2 mm., abd. 3 mm. 

Cephalothorax reddish brown ; abdomen brown folium covering two- 
thirds the width of abdomen, sides golden; legs light brown; mandibles, 
superior margin of fang groove one large hooked tooth near insertion 
of fang and two smaller teeth close together near median edge; inferior 
margin, two small teeth about opposite the two teeth on superior 
margin, fang long and overlapping; eyes, anterior row recurved, A.M.E. 
about a diameter apart and slightly larger than P.M.E. ; posterior row 
straight; lateral eyes touching; quadrangle of median eyes about square. 

Palpus; tibia and patella about equal length, bulb of palpal organ 
large, not divided, tarsus extending l)ut short distance beyond, tarsal 
appendage slender, with slender curved tip, with the usual widened 
portion one-third from tip. 

9 5.8 mm. long; ceph. 2.5 mm., abd. 3.5 mm. 

The colors and position of the eyes are the same as in the male. The 
mandibles are as Keyserling describes, an even convex margin when 
seen from the side, not suddenly geniculate as in P. xanthosiovia (brcvis 
Keys.). On the superior margin of the mandible are three small teeth 
corresponding to those in the male; the inferior margin also has three 
teeth opposite those on the superior margin. 

Allotype (d^); New York, Sea Cliff, N. Banks Coll. 

cf 9 N. Y.; Ithaca; D. C; Washington. 

Pachygnatha males 

1. Eyes of posterior row equidistant; superior margin of mandibles with two 

large sharp teeth near median edge; tarsal appendage of palpus long and 

slender; a small species autumnalis 

Eyes of posterior row not equidistant 2 

2. Mandibles with tooth above insertion of fang; superior margin with three 

teeth on median half; fang sinuate, with tooth about middle on inner 
edge; tarsal appendage of palpus with truncate tip; a small species, 

dorothea 
Mandibles without tooth above insertion of fang 3 



178 bulletin: museum of comparative zoology 

3. Fang with tooth about middle on inner edge; superior margin of mandible 

with a large tooth opposite tooth of fang and two small teeth near 

median edge; bulb of palpus about half length of tarsus tristriata 

Fang without tooth on inner edge 4 

4. Superior margin of mandible with three equidistant teeth, the basal 

largest; tarsal appendage of palpus broad and spatulate 

xanthostoma (brevis) 
Superior margin of mandible with three teeth, basal one large and sharp 
and two very small ones near median edge; apical half of tarsal appen- 
dage very slender, ending in a sharp point furcillata 

Gasteracantha cancriformis (Linne) 

Vibradellus carolinus ChamberUn, Bull. Mus. Comp. Zool., 1925, 67, p. 214. 

An examination of the type of J 'ibradclhis carolinus shows that it is a 
male Gasicracantha cancriformis. Unfortunately the specimen is faded 
and broken and the fourth pair of legs are missing so that the presence 
or absence of the comb characteristic of the Theridiidae can not be 
verified. In the description ]\Ir. Chamberlin fails to mention the row 
of sigillae on the basal margin of the abdomen, but calls attention to 
" a rounded shoulder or protuberance a little behind each anterolateral 
cornor" which is all that remains in the male of the large anterolateral 
spines in the female. The arrangement of eyes is characteristic of 
many males in the Epeiridae. 

THOMISIDAE 

Xysticus banksi nom, nov. 

Xysticus pallidus Bryant, Psyche, 1930, 37, p. 138, figs. 11, 12, 14, preoccupied 
by Cockerell, Ent. Month. Mag.; 1890, p. 191. 

Xysticus laticeps spec. nov. 
Plate 3, fig. 25 

9 6.5 mm. long; ceph. 3 mm. long, 3 mm. wide, head 2.5 mm. 
wide; abd. 3.8 mm. long. 

Cephalothorax longer than I femur. 

Cephalothorax light brown, with a wide marginal dark stripe and 
dark stripes from lateral eyes converging and ending in a pair of large 
dark spots above the abdomen; usual dark spot at end of thoracic 
groove small; abdomen brown with irregular cream color marks and 
paired dark spots; abdominal spines heavy: sternum light with irregu- 



BRYANT: NEW AND LITTLE KNOWN SPIDERS 179 

lar dark spots; coxa3 light with two dark dots; legs light brown with 
faint lines of a darker brown; IV fenuir and tibia with a pair of dark 
spots at tip; spines, I tibia, 7-5, no lateral, metatarsus, G-5, 2 lateral. 

Epigynum a simple cavity with a narrow septum at base. 

Holo'type (9) Ala.; Mobile, 2 August, 1930, \Y. S. Creighton. 

This species belongs near Xysticus Iucta7is Koch but differs as follows: 
head broader, dark stripes on cephalothorax converging, not parallel, 
greater number of spines on I tibia and the shape of the epigynum 
It differs from Xysticus graminis Emerton by the greater size, the con- 
verging dark stripes, much longer legs and the more numerous spines 
and the epigynum which is broader than long. 

Xysticus trimaculatus spec. nov. 
Plate 2, figs. 12, 13 

cf 4.5 mm. long; ceph. 2.5 mm., abd. 2.2 mm. 

Cephalothorax shorter than I femur. 

Cephalothorax golden brown with three subequal dark brown spots 
on posterior part, a white band between eye rows and a pair of white 
converging lines from posterior eye row to median dark spot, sides 
lightly veined with dark; long dark bristles at edge of clypeus, a pair of 
long bristles above each palpus and a row of smaller bristles outside the 
white line; abdomen white with many long dark bristles, one median 
dark spot at basal half followed by four pairs of irregular dark spots; 
sternum and coxpe light yellow; venter light; legs; I left leg missing, 
little lighter than cephalothorax, dark ring at tips of I and II tibiae, 
I and II metatarsi darker and a dark stripe beneath an apical half and a 
fringe of long white hairs on under side, longer than spines; III and IV 
legs with dark spots at tip of femora and tibiae; spines; I femur, a row 
of 6 long spines on upper edge, 5 shorter spines in front, many short 
dark hairs; I tibia 4-4, 3 lateral, metatarsus, 3-3, 2 lateral, all femora 
have a row of long dark spines on upper edge that are very conspicuous. 

Palpus; tibia shorter than patella, superior apophysis rather small, 
inferior apophysis with an upper lobe extending towards the palpal 
organ and a semicircular lobe with a distinct keel at right angles to the 
tibia, three long bristles on inner edge of tibia; tutaculum small, usual 
apophyses in center of palpal organ simple and widely separated. 

Holotype (d") .\rk.; Hope, 9 June 1931. INIiss L. Knobel Coll. 

This is closely related to the light form of Xysticvs triguttatvs but 
differs in the following points; the much larger size, the comparative 
length the first femur and the cephalothorax, the lighter color cephalo- 



180 bulletin: museum of comparative zoology 

thorax and the light legs with a dark line beneath first and second meta- 
tarsi, and in the palpus by the different shaped inferior apophysis of 
the tibia and in the palpal organ by the widely separated central 
apophyses. 

Xysticus variabilis Keyserling 
Plate 2, fig. 11; Plate 3, fig. 31 
Die Spinnen Amerikas, Laterigradae, 1880, p. 40, pi. 1, fig. 19. 

9 3 mm. long; ceph. 1.5 mm.; abd. 1.8 mm. 

Cephalothorax longer than I femur. 

Cephalothorax light brown with a median area blotched with cream 
color, sides darkened with irregular dark Aeins and a pair of large dark 
spots in front of the abdomen, usual dark spot at end of thoracic groove 
small; abdomen with a pair of dark spots at base and three or four dark 
transverse bars, broken by a median cream color branching figure; 
sternum and venter light with scattered dark marks; coxse and under 
side of legs light; legs light brown with cream color blotches, IV leg 
dark bands at tips of femur, patella and middle and tip of tibia; spines, 
I tibia, 4-3, no lateral, metatarsus, 3-3, all longer than diameter of 
joint, 1 lateral, at the edge of clypeus are several long bristles equal in 
length to space between A.M.E. and in eye area several long clavate 
bristles; abdomen with many small clavate bristles from coriaceous 
pits. 

Epigynum an oval depression, wider than high, with a dark median 
piece and two dark spots below as figured. 

The specimen agrees fairly well with Keyserling's description but is a 
little smaller. The clavate bristles in the eye area and on the abdomen 
are not mentioned. 

cf 3 mm. long; ceph. 1.7 mm., abd. 1.5 mm. 

Cephalothorax longer than I femur. 

Cephalothorax brown with a faint V shaped light mark ending at the 
thoracic groove, median line broken into five dark dashes ending with a 
rather small dark spot at thoracic groove, sides brown, veined with a 
darker brown and a pair of large dark spots just in front of the ab- 
domen, very few bristles on cephalothorax; abdomen lighter browTi 
with a pair of dark blotches at base and three broken dark bands across 
posterior half with parallel cream color lines above and irregular cream 
color lines on sides, bristles rather large; venter and under sides of legs 
light; legs light brown, I and II femora with dark spots, tibiae with 
indistinct dark band at base; III and IV legs dark bands at tips of all 



BRYANT: NEW AND LITTLE KNOWN SPIDERS 181 

joints; spines, I femur, 4 rather short on front side, 3 on upper edge, 
tibia 4-4, 2 middle pairs long, 3 lateral; metatarsus 3-3, 2 lateral, first 
and second pairs very long. 

Palpus; superior apophysis of tibia large, inferior apophysis long, 
ending in a curved knob that rests on the palpus ; tutaculum large and 
curving on the lower side of the palpus ; palpal organ with the usual two 
apophyses which do not stand upright but are turned over and some- 
what appressed to palpal organ, subequal with the superior somewhat 
sinuous; style ending in a circular loop with tip outside the palpus. 

Allotype (cf ) N. C; Wilmington, May 1900, J. H. Emerton Coll. 

1 9 N. C; Newbern, May 1900; 1 d" S. C; Charleston, May 1900; 
J. H. Emerton Coll. 

This species was described from a female in the Simon Collection. 
Mr. Banks, after seeing the type, states in Proc. Phila. Acad., 1913, p. 
179, "One female, a small species, which resembles a young stomacho- 
sus." While the males were not found with the female, the size and 
markings make it quite certain that the two belong together and the 
male palpus shows that it is a distinct species. 

Philodromus bilineatus spec. nov. 
Plate 2, figs. 14, 19; Plate 3, fig. 29 

9 4 mm. long; ceph. 1.5 mm., abd. 2.5 mm. 

Cephalothorax light brown with broad dark stripes extending from 
the P.L.E. to posterior edge of cephalothorax; a faint light lateral 
stripe; abdomen light brown with the dark stripes on the cephalothorax 
continued and converging at the apex, sides cream white; sternum, 
labium and venter almost white; legs light brown with anterior and 
posterior dark stripes on femora, patellae and tibiae of all legs but most 
distinct on the III and IV. II leg longest; spines I, tibia, 2-3, 1 lateral, 
metatarsus, 2-2, II tibia, 2-2-2-2, metatarsus, 2-2, 2 lateral; eyes, 
black surrounded by white rings; anterior row recurved, subequal, 
A.M.E. nearer A.L.E. than to each other; posterior row recurved, 
P.M.E. smallest of the eight and widely separated; quadrangle of 
median eyes much wider than high and once and a half wider behind 
than in front; lateral eyes on separate tubercles at the extreme edge of 
the carapace; clypeus more than twice the diameter of anterior eye with 
a few long bristles ; cephalothorax longer than wide with sides almost 
parallel; abdomen notched at base. 

Epigynum with a broad median lobe with two dark bodies each side 
as figured. 



182 bulletin: museum of comparative zoology 

cf 2.8 mm. long. 

Markings the same as on the female but less distinct and the stripes 
on the anterior legs reduced to rows of elongated dots. 

Palpus; tibia but little longer than patella with a stout, curved 
black hook which rests against tarsus, and an inferior, flat leaf-like 
lobe extending over the palpal organ. Palpal organ about filling tarsal 
cavity; the style is short and black and the tips rest against a trans- 
parent lobe in the upper part of the organ; loop is almost vertical. 

Holotype (cf) Fla.; Dunedin, March 1927, W. S. Blatchley Coll. 

Allotype (9) Fla.; Dunedin, March 1927. 

Paratype 9 Fla.; Royal Palm Park, March 1927. 

This species differs from Philodroni aides pruiariae Scheffer in the 
few spines under the anterior tibiae, but it agrees with it in the position 
of the lateral eyes on the extreme edge of the carapace. It agrees with 
Philodromus montanus spec. nov. in the general shape of the palpus, 
the two dark lines on the under side of the legs, but differs from it, by 
the arrangement of the eyes and the parallel sides of the carapace and 
the narrow and less sloping clypeus. 



Philodromus montanus spec. nov. 
Plate 2, fig. 20; Plate 3, fig. 26 

9 4 mm. long; ceph. 1.5 mm., abd. 2.5 mm. 

Cephalothorax with broad pale median stripe, sides darker with a 
narrow marginal light line; abdomen light, with vague transverse 
bands on apical half, sides darker; sternum light with a few dark dots 
about margin; venter light; legs, I missing, light with a dark line on 
anterior and posterior sides of femora, patellae and tibiae; spines II, 
tibia, 2-2-2, 2 lateral, metatarsus, 2-2, 2 lateral; eyes, anterior row 
recurved, A.M.E. slightly nearer A.L.E. than to each other; posterior 
row of eyes recurved, subequal, P.M.E. twice as far from each other as 
from P.L.E.; clypeus sloping and equal to space between A.M.E. 

Epigynum, median lobe narrow at l)ase with oval cavity each side as 
figured. 

cf 3 mm. long; ceph. 1.5 mm., abd. l.G mm. 

Cephalothorax brown, mottled with a darker shade and a little 
lighter in center; abdomen brown without any definite markings, but 
mottled with a darker Ijrown; sternum and venter pale yellow; legs 
light brown, almost yellow, with a dark line on anterior and posterior 
sides of femora, patellae, tibiae and metatarsi; spines, I tibia, 2-2-2, 



BRYANT: NEW AND LITTLE KNOWN SPIDERS 183 

2 lateral, metatarsus, 2-2, 3 lateral; eyes same as in female but clypeus 
is higher and more sloping. 

Palpus short, tibia shorter than patella, and tibia and patella not as 
long as tarsus; tibia with a short black hook on the superior side and a 
thin transparent leaf-like inferior apophysis which rests against the 
palpal organ; tarsal cavity almost filled with palpal organ, style starts 
above the middle and has a straight black tip which rests against a 
transparent lobe; the loop is oblique. 

Holotype (d") N. C; Black Mt.; North Fork Swannanoa River, 18- 
30 May, N. Banks Coll. 

Allotype (9 ) N. C; Black Mt. 

This pair was identified as Philodroinus rufus Walckenaer by Mr. 
Banks in his list of spiders from North Carolina. It belongs to the 
group of Philodromus with the P.M.E. widely separated and with dark 
lines on the legs, minidus Banks, inaequipes Banks, and carolinus 
Banks. It differs from minutus in the male palpus by the greater length 
of the inferior apophysis and the longer embolus, and from inaequipes 
by the narrower palpus and the shorter superior apophysis, 

Philodromus inaequipes Banks 
Plate 2, fig. 21 ; Plate 3, fig. 33 
Can. Ent., 1900, 32, p. 99. 

cf 3 mm. long; II femur 2.3 mm. long. 

Cephalothorax dark brown, with a very faint light V shaped mark 
over thoracic groove; abdomen dark brown with a broad median stripe 
of a darker brown; sternum, coxse and under side of legs light brown; 
venter white; legs light brown, I pair missing, a dark line on anterior 
and posterior sides of II femur, and anterior side of III and IV femora 
and tibiae; eyes, anterior row recurved, subequal, A.M.E. slightly 
nearer A.L.E. than to each other; posterior row recurved, P.M.E. 
separated by nearly twice the space between A.M.E. 

Palpus; patella and tibia about equal length, tibia as broad as long, 
the superior apophysis a black hook curved towards the palpus, equal 
in length to more than half the diameter of the joint; inferior apophysis 
a flattened white lobe which rests against the palpal organ; palpal 
organ fills the tarsal cavity; style short, starts about the middle and 
extends obliquely across the cavity; loop is broad and is parallel to 
embolus. 

Allotype {&) Va.; Falls Church, N. Banks Coll. 

1 9 Va.; Falls Church, N. Banks Coll. 



184 bulletin: museum of comparative zoology 

This species was described from a single female from Washington, 
D. C. Among the unidentified material in the museum, a male and 
female from Falls Church are this species. It belongs to the section of 
Philodromus with widely separated P.M.E. and dark lines on the sides 
of the legs. 

Philodromus emertoni spec, nov, 
Plate 2, figs. 17, 22; Plate 3, fig. 34 

cf 3.5 mm. long; ceph. 1.5 mm. 

Cephalothorax light brown, without median light stripe but with 
scattered dark spots on chpeus, around eyes and ahout dorsal groove; 
abdomen brown, without markings and covered with iridescent hairs; 
sternum light with dark spots about margin; coxse, venter and under 
side of legs light; legs light brown with scattered dark dots on all 
joints, II leg much longer than I leg; spines I femur, 3 on upper side, 
3 basal, tibia, 2-2-2, 3 lateral, all longer than diameter of the joint, 
but slender and colorless, metatarsus, 2-2-2; scopula thin; cephalo- 
thorax as long as wide; eyes, anterior row recurved, subequal and equi- 
distant; posterior row almost straight, P.M.E. separated by more than 
twice the distance between P.L.E. and P.M.E. 

Palpus rather short; tibia not as long as patella, superior apophysis 
short, acutely pointed with a small tooth on exterior side, inferior 
apophysis ^■ery close to superior, slender, thin and folded, extending on 
the upper side of the tarsus, so that in the ventral view only the tip is 
seen a little above the origin of the style; palpal organ fills the tarsal 
cavity, style starts on the exterior edge, extends across the middle of 
the upper part, making about two-thirds of a circle; the usual loop is 
almost horizontal. The small black spine sometimes found in the upper 
half just under the tip of the embolus is wanting. 

9 4.5 mm. long; ceph. 1.8 mm. 

Cephalothorax cream color with vague darker marks and dark dots 
on clypeus, about eyes and thoracic groove as in male; abdomen 
covered with small dark dots, but with a broad median lighter area 
which has four pairs of dark spots; sternum cream color with many 
dark dots about margin; coxte and venter light; legs light with many 
dark dots on all joints; spines as in male but longer and heavier; 
cephalothorax longer than wide; eyes as in male. 

p]pigynum has a broad median septum with a wide oval area each 
side. 

Holotype (cT) N. C; Newbern, May 1900, J. H. Emerton Coll. 



BRYANT: NEW AND LITTLE KNOWN SPIDERS 185 

Allotype (9 ) N. C; Newbern, May 1900, J. H. Emerton Coll. 

This species is similar to Philodromus ornatus Banks but the palpus 
is twice as large and much broader in proportion, the inferior apophysis 
is narrower and closer to the superior and the loop is nearer horizontal. 
The palpal organ is very similar to Pkilodromns robustus Emerton, but 
both superior and inferior apophysis are different and the legs of 
Philodromus robustus are very hairy. 

Philodromus aureolus (Clerck) 

Ara?}eus aureolus Clerck, Svensk. Spindl., 1757, p. 133, pi. 6, fig. 9. 
Philodromus canadensis Emerton, Can. Ent., 1917, 49, p. 270, fig. 22. 

Keyserling in Die Spinnen Amerikas, Laterigradae, 1880, states 
that this species has a wide distribution over the western states. On 
examining the type of Philodromus canadensis it proves to be the Euro- 
pean species. It is common throughout Canada and as far south as 
Boston, Massachusetts, and Ithaca, New York. It is possible that 
Philodromus californicus Keyserling may be this species. 

Philodromus infuscatus Keyserling 
Fig. 28 

Philodromus infuscatus Keyserling, Die Spinnen Amerikas, Laterigradae, 

1880, 1, p. 222, pi. 5, fig. 122,. 
Philodromus unicolor Banks, Proc. Phil. Acad., 1892, p. 61, pi. 3, fig. 22. 
Philodromus macrotarsus Emerton, Can. Ent., 1917, 49, p. 271, fig. 22, nos. 1, 2. 

This species was described by Keyserling from a female from Balti- 
more and by ]\Ir. Banks from a female from Ithaca. Mr. Emerton 
described the male from Vineland, Ontario. The male has been found 
since with females from Falls Church, Virginia, and from Poughkeepsie, 
New York. ]Mr. Emerton has an excellent figure of the male palpus. 

Philodromus laticeps Keyserling 

Philodromus laticeps Keyserling, Die Spinnen Amerikas, Laterigradae, 1880, 

p. 215, pi. 5, fig. lis. 
Philodromus louisianus Chamberlin, Proc. U. S. Nat. Mus., 1924, 63, p. 23, pi. 

5, fig. 39. 

The type of Philodromus louisianus Chamberlin proves to be 
Philodromus laticeps Keyserling, described from Georgia. The male 



186 bulletin: museum of comparative zoology 

is not known but the female has been found as far north as Newton, 
Massachusetts. 

Apollophanes texanus Banks 

Apollophanes texanus Banks, Jour. N. Y. Ent. Soc, 1904, 12, p. 113, pi. 5, fig. 

12, pi. 6, fig. 20. 
Philodromus syntheticus Chamberlin, Proc. Cal. Acad. Sci., 1925, 14, p. 124, 

figs. 33, 36. 

On examining the type of Philodromus syntheticus Chamberlin it is 
found to be Apollophanes texanus Banks. The strongly procurved 
posterior row of eyes which are equidistant, as well as the A.L.E., 
P.L.E. and P.M.E. which form an equilateral triangle, are characters 
of Apollophanes rather than Philodromus. 

Ebo inquisitor (Thorell) 

Plate 2, fig. 16; Plate 3, fig. 30 

Philodromus inquisitor Thorell, Bull. U. S. Geol. Surv., 1877, 3, p. 502. 
Philodromus thorellii Marx, Catalogue, 1889, p. 558, thorelUi preoccupied by 

Cambridge, 1872. 
Philodromus inquisitor Emerton, Trans. Conn. Acad., 1894, 9, p. 419, pi. 5, 

fig. 8. 

cf 4 mm. long, ceph. 1.6 mm. long, 1.8 mm. wide, abd. 2.4 mm. long, 
Cephalothorax with median mottled light stripe, sides dark brown; 
abdomen with spear-shaped dark mark at basal half, sides and apex 
dark; the contrast in color between basal half and apex not as great as 
in the female; sternum, coxae and venter light brown covered with 
minute brown dots; legs light brown covered with small brown dots 
which form bands at middle and tips of femora, base and tips of tibiae; 
I femur 2.5 mm. long, IT femur 4.5 mm. long; sternum a little wider 
than long; II coxse largest and IV coxte separated by a diameter; 
labium about as high as wide ; eyes, anterior row recurved so that upper 
margins are even, A.M.E. largest of the eight, and A.L.E. smallest, 
A.M.E. separated by a little more than a diameter and from the A.L.E. 
by about a radius; posterior row straight, subequal, P.M.E. separated 
by more than three diameters and from P.L.E. by about two diameters; 
quadrangle widest behind and about as high as wide between P.M.E.; 
clypeus vertical as high as quadrangle. 

Palpus, femur longer than cephalothorax, tibia longer than patella, 
tibia and patella not as long as femur, superior apophysis short, trun- 



BRYANT: NEW AND LITTLE KNOWN SPIDERS 187 

cate with distinct tooth at each corneras figured; inferior apophysis 
semitransparent, broad and a little shorter than superior; tarsus long 
and narrow, palpal organ filling cavity, style starts at upper third of 
organ and follows curve of cavity making a small semi-circle; loop 
small and horizontal. 

Allotype (cf) Cal.; Claremont, C.F.Baker, N. Banks Coll., many 
females from Colorado and Washington. 

By the characters used in separating genera, this species would be 
placed in Ebo rather than Philodromus. The large anterior median 
eyes, the posterior row of eyes almost straight and the second leg much 
longer than the first are all characters of Ebo, but the general appear- 
ance is not much like the type of the genus. Mr. Emerton states that 
he has seen the type of Philodromus inquisitor in the Packard Collec- 
tions and gives an excellent figure of a female from Laggan, Canada, 
which is now in the museum collection. He did not note the large an- 
terior median eyes or the greater length of the second leg. 

Ebo cockerelli spec. nov. 
Plate 2, figs. 15, 18; Plate 3, figs. 27, 36 

cf 3.5 mm. long; ceph. 1.5 mm. long, 1.6 mm. wide; abd. 2.5 mm. 
long. 

Cephalothorax light brown with a light V shaped mark ending at the 
thoracic groove; sides with dark veins; abdomen with a spear shaped 
dark mark at the basal third and indistinct chevrons formed by dark 
dots on apical half; sternum and coxae light brown with many dark 
dots; venter almost white; legs light yellow with dark bands formed by 
dots at tips of femora and base and tips of tibiae; spines small and in- 
conspicuous; spines, I tibia, 2-2, 2 lateral, metatarsus, 2-2, 1 lateral; 
thin scopula on I metatarsus and tarsus and on II, III and IV tarsi; 
II pair of legs much the longest, II femur twice as long as I femur; 
sternum one-third wider than long and widest between second coxse; 
labium about as high as wide at the base; IV coxffi separated by a di- 
ameter and II coxse slightly longest; eyes, anterior row slightly re- 
curved, A.M.E. largest of the eight, separated by more than a diameter, 
A.L.E. about a radius from A.M.E. and about half as large as A.M.E. ; 
posterior row slightly procurved, subequal, P.M.E. nearer P.L.E. than 
to each other; quadrangle not quite as high as space between P.M.E. ; 
clypeus vertical about equal to height of quadrangle. 

Palpus longer than cephalothorax; tibia nearly twice as long as 
patella and longer than tarsus; superior apophysis of tibia a short 



188 bulletin: museum of comparative zoology 

truncate lobe with three apical teeth; inferior apophysis a white leaf- 
like lobe folded against superior apophysis; palpal organ very simple; 
style starts in upper third and follows contour of the cavity, ending 
about opposite to its origin; loop oblique; the usual dark spine found in 
Philodromus aureolus missing. 

9 4 mm. long; ceph. 1.6 mm. long, 1.7 mm. wide; abd. 2.5mm. long. 

The colors and markings are the same as in the male but the chev- 
rons at the apex of the abdomen are not as distinct and are more like 
two converging dark lines; the dark dots on the legs are more scattered 
and the dark bands less conspicuous; the II pair of legs is much the 
longest and the II femur is once and a half the length of the I femur; 
eyes as in male. 

Epigynum has a broad septum with two obliciue dark lines on each 
side. 

Holotype (d") New Mexico; Mesilla Park, T. D. A. Cockerell; 
N. Banks Coll. 

Allotype (9 ) New Mexico; Mesilla Park. 

Paratype (9) Col.; Boulder, T. D. A. Cockerell. 

This species differs from Eho inquisitor (Thorell) in the male palpus 
and the epigynum. They both have Philodromus markings, but agree 
with the characters used to separate Eho from Philodromus. 



Ebo oblongus Simon 
Plate 3, fig. 23 
Ann. Soc. Ent. Belgique, 1895, 39, p. 442. 

9 6.3 mm. long; ceph. 2.5 mm. long, 2.5 mm. wide, abd. 3.8 mm, 

long. 

Cephalothorax with median light stripe, sides mottled with a dark 
brown ; abdomen light with median dark lanceolate mark on basal half 
followed by faint chevrons and irregular dark marks on sides of pos- 
terior half; legs light yellow, lighter than the cephalothorax with darker 
spots at base of spines and many minute tlots forming l)roken bands at 
middle and tips of femora and over the entire length of tibiae; sternum 
and venter light yellow covered with minute dark dots; coxffi yellow 
with few dots; clypeus almost vertical and as high as quadrangle of 
median eyes; labium wider than high; coxa^ subequal; sternum longer 
than wide, widest between second coxa? and prolonged in a blunt point 
between fourth coxai; fourth coxte se])arated by almost a diameter; 
eyes, anterior row recurved, A.M.E. largest of the eight and separated 



BRYANT: NEW AND LITTLE KNOWN SPIDERS 189 

by a diameter and a half, and by half a diameter from A.L.E.; posterior 
row but little longer than anterior, only slightly recurved, eyes sub- 
equal, P.M.E. slightly nearer P.L.E. than to each other; quadrangle as 
high as distance between P.M.E. ; A.L.E., P.M.E. and P.L.E. form an 
equilateral triangle; left legs missing, II leg longer than I; II femur one- 
fifth longer than I; spines, I and II til)iae, 2-2-2, 3 lateral, metatarsi, 
2-2-2, longer than diameter of joint, 3 lateral. 

Epigynum openings are a pair of transverse slits in a semicircular 
light area, above two pairs of round dark spots as figured. 

19 Ala.; Auburn, C. F. Baker, N. Banks Coll. 

Simon described this species from specimens about half grown. A 
co-type given by him to Mr. Banks is now in the museum collection. 
The adult specimen has been with the unidentified Pkilodromus. It 
agrees with Eho latithorax Keyserling in the large A.IVI.E., the short 
and nearly straight posterior row of eyes and the vertical clypeus, but 
it differs in the spiny legs and the second pair of legs which are longer 
than the first but are not greatly elongated, and the cephalothorax 
which is only as wide as long. 

TiBELLUS MARITIMUS (Mcnge) 

Plate 3, fig. 35; Plate 4, fig. 46 

Thanatus maritivms Menge, Preussische Spinnen, 1S74, p. 398, pi. 225. 
Tibellus oblongus Simon, Arachn. France, 1875, 2, p. 311, pi. 8, fig. 12. 
Tibellus maritimus Kulczynski, Mem. Acad. Imp. Sci. St. Petersburg, 1908 
(8), 18, p. 69-70. 

9 7.3 mm. long; ccph. 2.5 mm., abd. 5 mm. 

Cephalothorax pale yellowish brown, a broad median dark stripe 
with many short stiff hairs directed forwards, scattered short hairs 
directed forwards on the sides; abdomen bifid at base, pale brown with 
a narrow median dark stripe without pair of dark spots near apex; 
sternum, mouth parts and coxae pale yellow with dark hairs directed 
forward, venter pale brown; legs pale yellow, IV leg longest and much 
longer than III leg, scopulate on entire length of tarsi and metatarsi; 
labium longer than wide; sternum longer than wide in proportion of 
5 to 4; IV coxae almost touching; eyes, anterior row recurved, subequal 
and equidistant, posterior row very strongly recur^'ed, P.M.E. slightly 
smaller than P.L.E. and closer to each other than to P.L.E., quad- 
rangle of median eyes slightly higher than width between A.M.E.; 
clypeus about as high as quadrangle and almost vertical. 

Epigynum with a narrow median septum as figured. 



190 bulletin: museum of comparative zoology 

9 Alaska; Fox, 13 June 1931, Dr. G. Tullock. 

9 Wash.; Friday Harbor, 5 July 1927, L. G. Worley. 

It is very probable that this species is distributed across the northern 
part of America. Kulczynski states that from the figure by Emerton 
in Hentz, Spiders of North America, pi. 20, fig. 11, probably T. viari- 
timus, not T. duttonii which is supposed to be T. ohlongus, is figured. 
It is not surprising that the two species have been confused in America 
as European authors haA^e not always distinguished between them. 
According to de Lessert in Catalogue des Invertebres de la Suisse, 
1910, T. ohlongus always has a pair of dark spots near the tip of the 
abdomen in both sexes, which is wanting in T. maritimus. 

CLUBIONIDAE 
Myrmecotypus cubanus Banks 
Plate 4, figs. 43, 45 
Trans. Ent. Soc. London, 1926, 24, p. 433, fig. 1. 

9 4 mm. long; ceph. 1.8 mm.; abd. 2 mm. 

Cephalothorax brown, a little paler about eye area; abdomen brown 
with scattered white flattened hairs; sternum, mouth parts and I 
coxse brown, II, III and IV coxaj and trochanters white, venter brown 
with white hairs on sides; legs slender, I and II femora light l)rown, 
patellae, tibiae and metatarsi lighter, with dark lateral stripes, III and 
IV brown, femora with white hairs; spines, 2 long spines near apex of 
femora, I tibia, 2-2, metatarsus, 2-2; eyes; anterior row straight, 
A.M.E. largest of eight, a little more than a diameter apart and about 
a radius from A.L.E.; posterior row but little longer than anterior, 
straight and subequal, P.M.E. separated by about three diameters and 
from P.L.E. by at least two diameters; quadrangle almost square; 
clypeus with scattered long, white hairs about as high as quadrangle. 
Cephalothorax with a broad head, moderately high, widest between II 
and III legs, cephalic groove very short; abdomen with a scutum cover- 
ing more than basal two-thirds, widest at end of scutum ; large corneous 
plate on basal third of venter. 

Epigynum very simple, two oval openings widely separated. 

19 Texas; Brownsville, C. Scheffer. 

The genus Myrmecotyjms was made by O. P. Cambridge in 1894 for 
the s^iecies f^diginosvs. Simon in Hist. Nat. Araignees, 1897, 2, p. 175, 
places the genus as a synonym of A'pochinovima, Pavesi, F. O. P. Cam- 



BRYANT: NEW AND LITTLE KNOWN SPIDERS 191 

bridge in Biol. Centr. Amer., 1899, 2, p. 80, after examining a female 
Apochinoinvia sent him by Simon, shows that the two are distinct 
genera. Mr. Banks described two species from Panama in 1929, both 
differing from the three Mexican species. 

Trachelas bicolor Keyserling 
Verb. z. b. Ges. Wien, 1887, 37, p. 440, pi. 6, fig. 15. 

The type is in the museum collection and is from Haiti. Mr. W. S. 
Blatchlcy found a female at Royal Palm Park, Florida, 24 March 1925. 
It differs from Trachelas laticeps by the eyes and the epigynum. The 
male is not known. 

Trachelas laticeps spec. nov. 

Plate 3, fig. 24 

9 5 mm. 

Cephalothorax brown, rugose with many short hairs; abdomen pale 
yellow, covered with fine short hairs and scattered long bristles; 
sternum, labium and maxillae brown with scattered hairs; venter 
light; legs, I brown, enlarged, II, III and IV pale yellow, no spines but 

I tibia with a row of small cusps, I metatarsus with two rows of cusps, 

II metatarsus with one row of cusps, a dense brush of black hairs at tip 
on ventral side of tibiae III and IV; spur on posterior side of IV 
patella; III without any cusps on femur, patella or tibia; eyes, anterior 
row straight, subequal, A.M. E. separated by less than a diameter and as 
far from P.M.E., more than a diameter from A.L.E.; posterior row 
recurved and much longer than anterior row, P.M.E. about two di- 
ameters apart; clypeus not as high as diameter of A.M.E.; mandibles 
large and swollen, two teeth on inferior margin of fang groove and 
three teeth on superior margin. 

Epigynum as figured. 

Holotype (9 ) Fla.; Powelton P. O., Mrs. C. M. Willard. 

Co-types (9) Fla.; Royal Palm Park, 7-14 March 1930, C. W. 
Blatchley; 9 Fla.; Sebastian, April 1932, G. Nelson. 

Trachelas laticeps differs from T. tranquillus and T. bicolor by the 
epigynum and the slight differences in eye arrangement. In T. tran- 
quillus the openings of the epigynum are large, dark and are separated 
by less than a radius, and the A.M.E. are the largest; in T. bicolor the 
openings of the epigynum are small, about a diameter apart and the 
epigastric area is almost twice as high as wide, the eyes of the anterior 



192 bulletin: museum of comparative zoology 

row are subequal; in T. laiiceps the openings are much larger, are 
separated by almost a diameter and the area is almost twice as wide 
as high. The males of T. bicolor and T. laticrps are unknown. 

CoRiNNA GRACILIPES (Kcyserling) 

Plate 3, fig. 32 

Hypsinotus gracilipes, Keyserling, Verb. z. b. Ges. Wien; 1887, 37, p. 448, pi. 6, 
fig. 19. 

cf 9 mm. long; ceph. 4.5 mm. 

Cephalothorax dark red brown, rugose, with a few long hairs about 
eyes; abdomen light gray; sternum, maxillae and labium dark brown; 
venter pale; legs and palpi pale brown; spines, I tibia, 2-2-2-2-2-2, 
metatarsus, 2-2; II tibia, 5-3, metatarsus, 2-2; eyes, anterior row 
almost straight, A.M.E. largest, separated by a scant diameter and 
about a radius from A.L.E.; posterior row a little longer than anterior 
and slightly procurved, eyes equidistant, P.M.E. smallest of the eight; 
quadrangle wider than high; clypcus higher than diameter of A.lNI.Pl; 
mandibles swollen at base, porrect and transversely corrugated, four 
teeth on inferior margin of fang groove and three teeth on superior 
margin with a dense scopula of long hairs on the outer side. 

Palpus; tibia almost twice as long as patella with many processes 
near the tip as figured. 

Allotype (cf ) Fla.; Miami, July 1916. 

The type, a female, is from Haiti. This is undoubtedly the male as 
it agrees with the type in the museum collection in the eye arrange- 
ment and the number of spines on the anterior legs. 

PsEUDOSPARiANTHis CUBANA Banks 

Second Rep. Centr. Exper. Sta. Cuba, 1909, p. 165, pi. 45, fig. 4. 

Among the spiders collected by ]\Ir. W. S. Blatchley at Royal Palm 
Park, Florida, is a female of this species. It is the first time that the 
genus has been found in the United States. 

ATTIDAE 

IVIetaphidippus longipalpus Cambridge 

Biol. Centr. Amer., 1901, 2, p. 264, pi. 23, fig. 12. 

One male and two females were found by Mr. C. Schaeffer at Browns- 
ville, Texas. The first record of this Central American species in the 



BRYANT: NEW AND LITTLE KNOWN SPIDERS 193 

United States. The teeth on the mandible and the male palpus are 
very characteristic. 

Icius ciNCTiPES Banks 
Plate 4, figs. 42, 47 

Can. Ent., 1900, 32, p. 101. 

9 2.5 mm. long; ceph. 1 mm.; abd. 1.8 mm. 

Cephalothorax black about eyes, covered with short, white hairs, 
thoracic part brown witli a dark margin; abdomen dirty white with 
dark markings; sides dark, venter infuscate; legs light with dark spots 
on ventral sides of all joints, forming a broken ring at base of anterior 
tibiae and metatarsi; sternum light brown without marks; palpi with 
dark bands on upper side of patella and tibia, two terminal joints 
slightly enlarged; legs 4, 3, 1, 2, first pair of legs slightly enlarged, 
patella and tibia of equal length, metatarsus but little shorter than 
tibia and as long as tarsus; spines, I tibia, 2-2, metatarsus 2-2, longer 
than diameter of joint and overlapping; posterior legs almost devoid of 
spines; I coxae separated by less than their diameter; sternum oval, 
widest between II coxse; IV coxse separated by less than half a di- 
ameter; inferior margin of mandible with one tooth; eye area occupying 
about two-fifths of the cephalothorax ; anterior row of eyes recurved so 
that upper margins are even ; third row not quite as wide as cephalo- 
thorax and the small eyes nearer the first than the third row. 

Epigynum is a pair of simple oblique openings close together, not 
showing any of the internal structure. 

1 9 Fla.; Royal Palm Park, March 1930, W. S. Blatchley Coll. 

This species was described from females from Baton Rouge, Louisi- 
ana. An immature specimen from Punta Gorda, Florida, listed by 
Mr. Banks as Habrocestum pulex (Hentz) is undoubtedly this species. 
The generic position is uncertain. Most species of Icius are larger and 
have three pairs of spines beneath the first tibia and the cephalothorax 
is usually much lower and flatter than in tliis species, but until a male 
is found, it is thought best to leave it in the genus Icius which harbors 
many species that do not belong to it. 



EXPLANATION OF PLATES 



PLATE 1 



Bryant — New and Little Known Spiders. 



PLATE 1 

Fig. 1. Euryopis emertoni Bryant, ventral view of left palpus. 

Fig. 2. Euryopis ornata Bryant, dorsal view of male. 

Fig. 3. Euryopis ornata Bryant, ventral view of left palpus. 

Fig. 4. Theridion quadrimaculatum (Banks), ventral view of left palpus. 

Fig. 5. Gonatium crassipalpus Bryant, epigynum. 

Fig. 6. Gonatium crassipalpus Bryant, dorsal view of left palpus. 

Fig. 7. Dipoena lineatipes Bryant, epigynum. 

Fig. 8. Gonatium crassipalpus Bryant, lateral view of male cephalothorax; 

Fig. 9. Gonatium crassipalpus Bryant, lateral view of left palpus; 

Fig. 10. Gonatium crassipalpus Bryant, first leg of male. 



BULL. MUS. COMP. ZOOL. 



Bryant. New Spiders. Plate 1 







N; 



.^ 



-r^' 



V 



10 



PLATE 2 



Bryant — New and Little Known Spiders , 



PLATE 2 

Fig. 11. Xysticus variabilis Keys., ventral view of left palpus. 

Fig. 12. Xysticus trimaculatus Bryant, lateral view of left palpus. 

Fig. 13. Xysticus trimaculatus Bryant, ventral view of left palpus. 

Fig. 14. Philodromus bilineatus Bryant, eyes of female. 

Fig. 15. Ebo cockerelli Bryant, ventral view of left palpus. 

Fig. 16. Ebo inquisitor (Thorell), tibial apophysis of left palpus. 

Fig. 17. Philodromus emertoni Bryant, tibial apophysis of left palpus. 

Fig. 18. Ebo cockerelli Bryant, tibial apophysis of left palpus. 

Fig. 19. Philodromus bilineatus Bryant, ventral view of left palpus. 

Fig. 20. Philodromus montanus Bryant, ventral view of left palpus. 

Fig. 21. Philodromus inaequipes Banks, ventral view of left palpus. 

Fig. 22. Philodromus emertoni Bryant, ventral view of left palpus. 



BULL. MUS. COMP. ZOOL. 



Bryant. New Spiders. Plate 2 







PLATE 3 



Bryant — New and Little Known Spiders. 



PLATE 3 

Fig. 23. Ebo oblongus Simon, epigynum. 

Fig. 24. Trachelas laticeps Bryant, epigynum. 

Fig. 25. Xysticus laticeps Bryant, epigynum. 

Fig. 26. Philodromus montanus Bryant, epigynum. 

Fig. 27. Ebo cockerelli Bryant, eyes of male. 

Fig. 28. Philodromus infuscatus Keys., epigynum. 

Fig. 29. Philodromus bilineatus Bryant, epigynum. 

Fig. 30. Ebo inquisitor (Thorell), eyes of male. 

Fig. 31. Xysticus variabilis Keys., epigynum. 

Fig. 32. Corinna gracilepes (Keys.), lateral view of tibia of left palpus. 

Fig. 33. Philodromus inaequipes Banks, epigyniun. 

Fig. 34. Philodromus emertoni Bryant, epigynum. 

Fig. 35. Tibellus maritimus (Menge), epigynum. 

Fig. 36. Ebo cockerelli sp. n., epigynum. 



BULL. MUS. COMP. ZOOL. 



Bryant. New SproERs. Plate 3 







/^ 




36 



PLATE 4 



Bryant — New and Little Known Spiders. 



PLATE 4 

Fig. 37. Pachygnatha doroihea McCook, lateral view of left palpus. 

Fig. 38. Pachygnatha furcillata Keys., ventral view of left mandible. 

Fig. 39. Pachygnatha furcillata Keys., lateral view of left palpus. 

Fig. 40. Pachygnatha autumnalis Keys., lateral view of left palpus. 

Fig. 41. Pachygnatha dorothea McCook, ventral view of left mandible. 

Fig. 42. Ictus cinctipes Banks, epigynum. 

Fig. 43. Myrmecotypus cubanus, Banks, epigynum. 

Fig. 44. Pachygnatha autumnalis Keys., ventral view of palpus. 

Fig. 45. Myrmecotypus cubanus Banks, dorsal view of female. 

Fig. 46. Tibellus maritimus (Menge), lateral view of left palpus. 

Fig. 47. Icius cinctipes Banks, dorsal view of female. 



BULL. MUS. COM P. ZOOL. 



Bryant. New Spiders. Plate 4 




Nuv 8 1933 



^1^1 



Bulletin of the Museum of Comparative Zoology 

AT HARVARD COLLEGE 

Vol. LXXIV, No. 7 



REPORTS ON THE SCIENTIFIC RESULTS OF AN 

EXPEDITION TO THE SOUTHWESTERN HIGHLANDS 

OF TANGANYIKA TERRITORY 

VII 

HERPETOLOGY 



By Arthur Loveridge 



With Three Plates 



CAMBRIDGE, MASS., U.S.A. 

PRINTED FOR THE MUSEUM 

October, 1933 



PUBLICATIONS 

OF THE 

MUSEUM OF COMPARATIVE ZOOLOGY 
AT HARVARD COLI>EGE 



There have been published of the Bulletin Vols. I to LXV, LXVII- 
LXXIV, of the Memoirs Vols. I to LI, LIV. 

The Bulletin and Memoirs are devoted to the publication of 
original work by the Officers of the Museum, of investigations carried 
on by students and others in the different Laboratories of Natural 
History, and of work by specialists based upon the Museum Collec- 
tions and Exploration. 

These publications are issued in numbers at irregular intervals. 
Each number of the Bulletin and of the Memoirs is sold separately. 
A price list of the publications of the Museum will be sent on 
application to the Director of the Museum of Comparative Zoology, 
Cambridge, Massachusetts. 



Bulletin of the Museum of Comparative Zoology 

AT HARVARD COLLEGE 

Vol. LXXIV, No. 7 



REPORTS ON THE SCIENTIFIC RESULTS OF AN 

EXPEDITION TO THE SOUTHWESTERN HIGHLANDS 

OF TANGANYIKA TERRITORY 

VII 

HERPETOLOGY 



By Arthur Loveridge 



With Three Plates 



CAJSIBRIDGE, MASS., U. S. A. 

PRINTED FOR THE MUSEUM 

October, 1933 



Xo. 7. — Reports on the Seientific Results of an Expedition to 
the Southu'estern Highlands of Tanganyika Territory 

VII 

Herpetology 
By Arthur Loveridge 

The material discussed in the following pages was collected by the 
author with a view to throwing light on the faunal distribution of the 
southwestern highlands of Tanganyika Territory; this aspect of the 
collection is dealt with in the introduction to this series of reports 
where full information will be found as to altitudes and localities. 
The investigation was undertaken on behalf of the Museum of Com- 
parative Zoology and in part financed by a grant from the Carnegie 
Institute of Washington. 

Pt. I. Reptilia 
MATERIAL 

The period of collecting was from November 1, 1929 to July 9, 
1930, during which time 2,117 reptiles representing 125 species were 
preserved. This total comprised 1 species of crocodile, 7 of tortoises 
and turtles, 54 of snakes, 47 of lizards and 16 kinds of chameleon; in 
all 30 forms of reptiles were new to the collection of the Museum of 
Comparative Zoology. 

One might single out for special mention such rarities as : Chilorhino- 
phis gerardi, Rhinocalamus dimidiahis and Vipera superciliaris among 
snakes; Paragonatodes quattuorseriatvs, Chamaesaura miopropvs and 
Ablepharus megalurus among lizards and Chameleon anchietae, Cham- 
aeleon fullehorni and Brookesia platyceps in the Rhiptoglossa. 

Naturally as thorough a study of this material as has been possible 
involves quite a number of taxonomic alterations. Nor is this re- 
markable, for in planning the itinerary I arranged to visit the type 
localities of a dozen species of questionable status with the object of 
securing adequate series of topotypic specimens so as to elucidate 
their relationships and range of variation. 



198 



bulletin: museum of comparative zoology 



SUMMARY OF TAXONOMIC ALTERATIONS 

The following new species or races from this collection have already 
been described briefly; additional information regarding them will be 
found in the present paper. 



Lycophidion capense uzungwensis 
Athens barbouri 
Agama agama turuensis 

Agama agama ufipae 
Zonurus ukingensis 
Amphisbaena mpwapwaensis 
Chamaeleon werneri dabagae 
Chamaeleon incornutus 

Chamaeleon laterispinis 



Dabaga & Kigogo, Uzungwe Mtns. 
Dabaga & Madehani, Ukinga Mtns. 
Unyanganyi, Turu & Mangasini, 

Usandawi. 
Kipili, Ufipa on Lake Tanganyika. 
Tandala, Ukinga Mtns. 
Mpwapwa, Ugogo. 
Dabaga, Uzungwe Mtns. 
Madehani, Nyamwanga & Nkuka 

Forest, Rungwe. 
Kigogo, Uzungwe Mtns. 



In addition to the new species, the following are recorded from 
Tanganyika Territory for the first time: 

Pelusios nigricans rhodesianus Hewitt of Northern Rhodesia 

Typhlops graueri Sternfeld of Belgian Ruanda 

Philothamnus semivariegatus dorsalis (Bocage) of Angola 

Rhamphiophis acutus (Giinther) of Angola 

Vipera superciliaris Peters of Mozambique 

Atradaspis aterrima Giinther of West Africa, already known from Uganda 

Paragonatodes quattuorseriatus (Sternfeld) of Belgian Ruanda 

Lygodactylus piduratus gutteralis (Bocage) of Portuguese Guinea 

Lygodadylus angularis Giinther of Nyasaland 

Ichnotropis bivittata Bocage of Angola 

Chameleon anchietae Bocage of Angola 

Brookesia platyceps (Giinther) of Nyasaland 

while the undermentioned are revived: 

Homalosoma shiranum Boulenger as a race of Duberria lutrix (Linnaeus) 
Rhamphiophis rostratus Peters for East African "R. oxyrhynchus (Reinhardt)' ' 
Lygodactylus angularis Giinther sunk in L. piduratus (Peters) by Tornier 
Sepacontias modestus Giinther as a race of Riopa sundevallii (Smith) 

Certain reptiles, hitherto regarded as full species, are accorded 
subspecific rank, thus : 

Typhlops excentricus Procter as Typhlops schlegelii excentricus Procter 
Lycophidiuni acutirostre Giinther as Lycophidion capense acutirostre Giinther 
Leptophis dorsalis Bocage as Philothamnus semivariegatus dorsalis (Bocage) 



loveridge: African herpetology 199 

Hoiualosotua shiranum Boulenger as Duberria lutrix shiraniim (Boulenger) 
Boulengerina stormsi Dollo as Boulengerina annulata stormsi DoUo 
Nucras kilosae Loveridge as Nucras boulengeri kilosae Loveridge 
Sepaconiias modestus Giinther as Riopa sundevallii modestum (Giinther) 

The following are considered strict synonyms : 

Typhlops tornieri Sternfeld = Typhlops pundatus pundatus (Leach) 

*Typhlops humbo Bocage = Typhlops schlegelii mucruso (Peters) 

*Typhlops mandensis Stejneger =Typhlops schlegelii mucruso (Peters) 

Typhlops opisthopachys Werner = Typhlops pinguis Waite of Australia 

*Glauconia merkeri Werner = Leptotyphlops conjunda (Jan) 

Glaucoma latirostris Sternfeld = Leptotyphlops conjunda (Jan) 

*Glauconia disfanti Boulenger = Leptotyphlops scutifrons (Peters) 

Gastropyxis orientalis Werner =Hapsidophrys lineata Fischer 

*Prosymna variabilis Werner = Prosymna ambigua Bocage 

Amplorhinus taeniatus Sternfeld = Hemirhagerrhis kelleri Boettger 

*Rhamphiophis connali Parker = Rhatnphiophis oxyrhynchus (Reinhardt) 

Parkerophis Barbour & Amaral 1927 = Chilorhinophis Werner 1908 (1907) 

*Atradaspis phillipsi Barboui' = Atradaspis microlepidota Giinther 

Atradaspis magretti Scortecci =Atractaspis microlepidota Giinther 

*Lygodactylus manni Loveridge = Lygodadylus p. piduratus (Peters) 

Elasmodadylus triedrus Boulenger = Pachydactylus boulengeri Tornier 

Nucras emini Boulenger = Nucras b. boulengeri Neumann 

Melanoseps ater longicauda Tornier = Melanoseps ater (Giinther) 

*Chamaeleon tempeli wolffi Tornier =Chamaeleon tempeli Tornier 

Rhampholeon Giinther 1874 =Brookesia Gray 1864 

In addition to those species now considered synonyms, the under- 
mentioned should be removed from the East African list : 
Sternfeld and Loveridge's records of Typhlops dinga { = T.s. schlegelii) 

for Tanganyika Territory. 
Boulenger, Stejneger and Loveridge's records of Typhlops schlegelii 

for Uganda, Kenya and Tanganyika respectively. 
Lonnberg's record of Glaucoma scutifrons for Tanganyika Territory. 
Loveridge's record for Lcptofyphlops disfcniti for Tanganyika Territory. 
All records of Rhamphiophis oxyrhynchus for which substitute R. 

rostratus. 
All records of Atractaspis rosfrata for which substitute A. bibronii of 

which it is a synonym. 
All records of Agama hispida distanii for which substitute .4. h. armaia. 
Nieden's record of Latastia siebenrocki for Kenya and Tanganyika. 
Loveridge's records of Lygosoma ferrandi for Tanganyika Territory. 

*Type or paratype examined. 



200 bulletin: museum of comparative zoology 

ACKNOWLEDGEMENTS 

I take this opportunity of expressing my great indebtedness to Mr. 
H. W. Parker (British Museum) and to Mr. V. FitzSimons (Transvaal 
Museum) for answering various queries involving the examination of 
long series of specimens and Mr. Karl P. Schmidt for examining 
various types in the Berlin Museum collection; also IVIons. F. Angel 
(Paris Museum), Dr. F. Gaston de Witte (Congo Museum), Dr. 
Wilhelm Gotz (Stuttgart Museum) and Mrs. H. T. Gaige (University 
of Michigan Museum of Zoology) for loaning specimens or affording 
me facilities for their examination. Without this generous cooperation 
it would have been infinitely more difficult to arri^'e at reasonable 
decisions. 

List of Species Collected* 

Order Loricata page 

Family CROCODYLIDAE 204 

Crocodyliis niloticus Laurenti 204 

Order Testudinata 

Family TESTUDINIDAE 200 

Testudo yardcdis Bell 206 

* Testudo tornieri Siebenrock 20() 

Family CHELONIIDAE 207 

Eretmochelys imbricata (Linnaeus) 207 

Family PELOMEDUSIDAE 208 

Pelusios sinuatus (Smith) 208 

*Pelusios nigricans nigricans (Dondorft") 209 

Pelusios nigricans rhodesianu3 Hewitt 210 

*Pelomedusa galeata (Schoepff) 211 

Order Squamata 
Suborder Ophidia 

Family TYPHLOPIDAE 212 

Typhlops graueri Sternfeld 212 

*Typhlops punctatus punctatus (Leach) 213 

{Typhlopa schlegclii schlegelii Bianconi) 214 

*Typhlo2}s schlegelii viucruso (Peters) 216 

* TypJdops schlegelii excentricus Procter 220 

{Tiiphlops opisthopachiis Werner) 222 

Family LEPTOTYPHLOPIDAE 223 

Leptotyphlops emini (Boulenger) 223 

*An asterisk opposite a species indicates that examples are available for exchange. Species 
in parentheses are discussed though not collected. 



loveridge: African herpetology 201 

PAGE 

{LeptohjiMops longicauda (Peters)) 223 

Leptotyphlops conjuncta (Jan) 224 

Family BOIDAE 226 

Python sebae (Gmelin) ' 226 

Family COLUBRIDAE 231 

Subfamily Colubriiiae 

*Nairi.v olivaceus (Peters) 231 

Glypholycus bicolor Giinther 232 

*Boaedo7i lineatus Dumeril and Bibron 232 

*Lycophidion capens'e capense (Smith) 233 

Lycophidion intermediates between capense and acuti- 

roslrc Giinther 234 

Lycophidion capense uzungwensis Loveridge 235 

Pseudaspis cana (Linnaeus) 235 

Chlorophis emini (Giinther) 236 

*Chlorophis hoplogaster (Giinther) 236 

Chlorophis neglectus (Peters) 237 

*Philothamn'ns semivariegaius semivariegatus Smith 237 

*Philothamnus semivariegatus dorsalis (Bocage) 238 

(Hapsidophrys lineata Fischer) 239 

Coronella semiornata Peters 240 

Grayia thollo7ii Mocquard 240 

*Duberria lutrix shiranuru (Boulenger) 241 

Prosyvina amhigua Bocage 244 

Subfamily Dasypeltinae 245 

*Dasypeltis scaber (Linnaeus) 245 

Subfamily Boiginae 246 

Tarbophis semiannulatus (Smith) 246 

*Crotaphoj)eltis hotamboeia hotamboeia (Laurenti) 247 

*Crotaphopeltis hotamboeia tornieri (Werner) 248 

Amplorhinus yiototaenia (Giinther) 250 

*Trimerorhinus tritaeniatus tritaeniatus (Giinther) 250 

Rhamphiophis acutus (Giinther) 252 

{Rhamphiophis oxyrhynchus (Reinhardt)) 252 

* Rhamphiophis rostratus Peters 253 

Dromophis lineatus (Dumeril and Bibron) 254 

*Psammophis subtaeniatus Peters 254 

*Psammophis sibilans (Linnaeus) 255 

*Psammojjhis biseriatus Peters 256 

*An asterisk opposite a species indicates that examples are available for exchange. Species 
iu parentheses are discussed though not collected. 



202 bulletin: museum of compakative zoology 

PAGE 

Psammophis angolensis (Bocage) 257 

Thelotornis kirtlandii (Hallowell) 257 

*Dispholidus typus (Smith) 258 

Calamelaps unicolor (Reinhardt) 260 

Rhinocalamus diviidiatu^ Giinther 261 

Miodon gahoncnsis (Dumeril) 261 

ChiJorhinophis gerardi (Boulenger) 262 

Subfamily Elapinae 263 

Boulcngcrina amudata stormsi Dollo 263 

*Naja melanoleii.ca Hallowell 270 

*Naja nigricollis Reinhardt 271 

(Dendraspis augusticeps Smith) 273 

Family VIPERIDAE 273 

*Causus rhombeatus (Lichtenstein) 273 

Causus rcsimus (Peters) 274 

Caucus defilippii (Jan) 274 

Viper a superciliaris Peters 275 

*BiHs arietans (Merrem) 276 

Athcris barbouri Loveridge 277 

Atradaspis irregularis (Reinhardt) 278 

(Atractaspis conradsi Sternfeld) 279 

(Atractaspis bibronii Smith) 280 

{Atractaspis aterrima Giinther) 281 

{Atractaspis microlepidota Giinther) 281 

Suborder Lacertilia 

Family GEKKONIDAE 282 

Paragonatodes quattuorscriatus (Sternfeld) 282 

*Hemidactylus mabouia (Moreau de Jonnes) 283 

*Hemidactylus pcrsimiJis Barbour & Loveridge 284 

Hcviidactylus tropidolepis squamuJatus Tornier 284 

Hemidactylus werneri icerneri Tornier 285 

{Hemidactylus icerneri alluaudi Angel) 285 

*Hemidactylus brookii Gray 286 

Lygodactylus capensis capcnsis (Smith) 286 

Lygodactylus stepensoni Hewitt 287 

*Lygodactylus grotei Sternfeld 287 

^Lygodactylus picturatus picturatus (Peters) 288 

* Lygodactylus picturatus var. on Mombasa Id 289 

*Lygodactylus picturatus var. on Ukerewe Id 290 

*An asterisk opposite a species indicates that examples are available for exchange. Species 
in parentheses are discussed though not collected. 



loveridge: African herpetology 203 

PAGE 

*LygodactyJus jyicturotus (lufturalis (Bocage) 291 

*Ly(jodactylus angularis Giinther 292 

*Pac}iydacfylus houlcngcri Tornier 293 

(Phelsuma laticauda (Boettger) ) 295 

Family AGAMIDAE 296 

*Aga))!a hispida armaia Peters 29(5 

*Agama agama Honotus Boulenger 297 

*Agama agama mwanzae Loveridge 298 

* Agama agama turuensis Loveridge 299 

* Agama agama dodomae Loveridge 300 

* Agama agama vfipac Loveridge 300 

* Agama atricollis Smith 300 

Family ZONURIDAE 301 

Zonvrus ukingensis Loveridge 301 

Chamaesaura miopropus Boulenger 302 

Family VARANIDAE 303 

J'aranus mloHcus (Linnaeus) 303 

Family AMPHISBAENID.AE 304 

Amphishaena mpxcapicaensis Loveridge 304 

Family LACERTIDAE 304 

*Niwras houlengeri hoidcngeri Neumann 304 

*Latastia johnstonii Boulenger 306 

*Latastia longicaudata rcimli (Vaillant) 30S 

Ichnotropis hivittaia Bocage 308 

Ichnotropis squamulosa Peters 309 

*Eremias spekii spekii Giinther 310 

Family GERRHOSAURIDAE 311 

(Gerrhosaurus major major Dumeril) 311 

*Gerrhosaurus major zcchi Tornier 311 

*Gerrhosaurus fiavigidaris flavigularis Wiegmann 312 

Family SCINCIDAE 312 

*Mabuya maculilabris (Gray) 312 

Mahuya planifrons (Peters) 315 

*Mahtiya megalura (Peters) 316 

*Mabuya varia varia (Peters) 316 

*Mabuya striata (Peters) 319 

Riopa fernandi (Burton) 320 

*Riopa sundevallii simdevallii (Smith) 320 

*Riopa sundevallii modestum. (Giinther) 322 

*An asterisk opposite a species indicates that examples are available for exchange. Species 
in parentheses are discussed though not collected. 



204 bulletin: museum of comparative zoology 

PAGE 

*Ablephanis houtonii africanus Sternfeld 323 

*Ahlepharus waklbergii (Smith) 324 

Ahlcpharus megcdurus Nieden 325 

Melanoseps ater (Giinther) 326 

Family ANELYTROPIDAE 328 

Feylinia currori elegans (Hallowell) 328 

Suborder Rhiptoglossa 

Family CH.\IVIAELEONTID.\E 329 

Chamaeleon gracilis gracilis Hallowell 330 

*Chamaeleon dilepis roperi Boulenger 330 

*Chamaeleon dilepis quilensis Bocage 331 

*Chamaeleon dilepis dilepis Leach 332 

*Chamaclcon hitaeniatus hiiacniatus Fischer 333 

*Chaviaeleon hitaeniatus hohnclii Steindachner 333 

Chamaeleon anchietac Bocage 333 

^Chamaeleon goetzei Tornier 334 

*Chamacleon tempeli Tornier 335 

*Chamaeleon fiilleborni Tornier 338 

*Chamaeleon xccrneri werneri Tornier 338 

Chamaeleon tverneri dabagae Loveridge 339 

*Chamaeleon jacksoni vauerescecae Tornier 340 

*Chamaeleon incornutn.s Loveridge 340 

Chamaeleon laierispinis Loveridge 341 

(Brookesia temporalis (Matschie)) 342 

^Brookcsia platyceps (Giiiither) 343 



*i 



Systematic List of Species Collected 

CROCODYLIDAE 

Crocodylus niloticus Laurenti 

Crocodylus niloticus Laurenti (part), 1768, Syn. Rept., p. 53: "Habitat in 
India orientali, et Aegypto." 

Skin and skull (AI. C. Z. 30000) Mwaya, Lake Nyasa. 6. iii. 30. 

Distribution. Crocodiles were also seen on an affluent of the Ruvu 
River close to Bagamoyo; two were observed lazily swimming on the 
surface of Lake Tanganyika in Nj^amkolobay just after sun-up. ^Ihey 
remained for a couple of hours. One was seen at Ukerewe Island and 
very many on Lake Victoria just above the Ripon Falls. 

*\n asterisk opposite a species indicates that examples are available for exchange. Species 
ill parentheses are discussed though not collected. 



loveridge: African herpetology 205 

Native name. Mamba (Kiswahili); ngivina (Kinyakusa). 

Measurements. The female listed above measured ten feet four 
inches (Head and body 1470 mm., tail 1400 mm., hindfoot 180 mm.). 

Diet, etc. On three occasions we crossed and recrossed the Mbaka 
River in an unusually crazy dugout tree-trunk. In fact during the ten 
days spent at Mwaya it was a matter of daily occurrence to cross 
some river or other but the dugout available on the ]Mbaka River 
was an exceptionally unstable and leaky affair with never less than 
three or four inches of water in the bottom. One day this dugout 
capsized both on the outward and return journey and shot my " boys" 
into the water. Salimu was highly incensed for, in addition to getting 
the cartridges wet, he lost some money and other belongings which 
were in his pocket at the time. He said that the ferryman told him 
that never a day passed without the dugout being upset. On the 5th 
of March I crossed at 7 a.m. and returned about 2 p.m., I was in- 
formed that crocodiles had taken two women at the crossing during 
the interval I had been away. The unfortunate women had waded 
into the water to fill their water pots. Turning to the chief's son, who 
accompanied me, I asked him how many people were taken in his 
district by crocodiles each month; he replied that about five were 
killed in this way and that they were mostly women engaged in draw- 
ing water. I asked why, seeing that bamboo was abundant, they did 
not follow the custom of intelligent natives such as those in the Moro- 
goro district and bail the water from a distance by means of a gourd 
attached to the end of a bamboo. "Too much trouble," he replied. I 
said I would shoot any crocodiles that I could if his people would 
come and tell me when they saw one basking. At 1 p.m. the following 
day I was summoned to shoot a crocodile that was lying, mostly 
concealed by grass in shallow water on the further side of the ri\'er. 
I could just see the top and back of its head and the first shot, a .351 
soft -nose bullet, right between the eyes was instantaneous in effect; 
a violent lashing of the tail accompanied by quivering of the limbs and 
in a minute all was still though there was an ever widening red patch 
on the water. The stomach held nothing but pebbles and sand. 

In strange contrast to the voracity of the crocodiles at Mwaya is 
their apparent indifference to the natives at both Nyamkolo and 
Kasanga near the southern end of Lake Tanganyika. At both places 
the natives, particularly the children, bathed freely near the shore. 
In answer to my enquiries nobody seemed to recall a case of a crocodile 
carrying off a person. Nevertheless my native personnel refused to 
bathe, preferring to bail out water and perform their ablutions at a 



206 bulletin: museum of comparative zoology 

safe distance. ^Yhen I invited Salimu's attention to the fact that 
little children were bathing with impunity though he and his com- 
panions would not, he replied that, " In many houses you may have 
noticed that a dog and cat will live together in amity but if a strange 
dog or cat appears the results may be different. These people and 
these crocodiles know each other of old but I do not know these 
crocodiles nor they me." 

Parasites. Nematodes were found in the body cavity of the Mwaya 
crocodile. 



TESTUDINIDAE 
Testudo pardalis Bell 

Testudo pardalis Bell, 1828, Zool. Journ., 3, p. 420: Africa. 

1 (M. C. Z. 30001) Saranda, Ugogo. 17. xii. 29. 
1 (M. C. Z. 30002) Tukuyu, Rungwe. 13. iii. 30. 

Distrihifion. A tortoise was described to me as occurring on Ukerewe 
Island which could be none other than the Leopard Tortoise though I 
failed to secure one during my brief stay. It has been recorded by 
Sternfeld. 

Native name. Malugangi (Kigogo). 

Habitat. The halfgrown male from Saranda was found walking 
about in the rather sandy thorn-bush country just before sunset. At 
the time the scanty grass was very dry after a prolonged drought. 

Did. Two very large examples from Kabete were given to me by 
Miss Gladys Leakey to take home to the London Zoological Society's 
gardens. Hitherto they had been fed on the foliage of ground nuts. 
On board I fed them every second day on lettuce and moistened soft 
bread of which they consumed quantities and it was rarely that they 
failed to clear up every leaf and cruml). Once or twice during the hot 
weather in the Red Sea they were soaked for an hour or so in a bath 
tub at which times the}- would drink deeply; they arrived in fine 
condition. 

Testudo tornieri Siebenrock 

Testudo tornieri Siebenrock, 1903, Ak. Wiss. Wien, Math.-nat. Klasse, 24, 
p. 185: "Bussisia" i.e. Busisi, Tanganyika Territory. Loveridge, 1928, 
Proc. U. S. Nat. Mus., 73, Art. 17, p. 49; Tabora; Dodoma; Mfilima; 
Kikombo; Kibakwe; Kondoa Irangi, Tanganyika Territory. 



loveridge: africax herpetology 207 

Testudo (Malacochersus) tornieri Lindholm, 1929, Zool. Anz. Leipzig, 81. p. 285. 
Tesiudo loveridgii, E. G. Boulenger, 1920, Proc. Zool. Soc. London, pp. 190-1. 

8 (M. C. Z. 30003-10) Mangasini, Usandawi. 13-16. xii. 29. 

Disiribution. The locality is a new one for the Soft-shelled Land 
Tortoise. 

Variation. In general tortoises in this series present a fairh' normal 
condition; all have four costals on either side of the five vertebrals 
and all but one have the normal eleven pairs of marginals, the ex- 
ception (M. C. Z. 30008) has twelve pairs. Two tortoises (M. C. Z. 
30003, 30009), however, displa^^ a striking variation of the nuchal 
shield which is not merely completely divided on the longitudinal 
axis but the two halves are separated b}' a wide V-shaped cleft. The 
description of the species should read, therefore, "a single nuchal, 
more rarely two." No. 30009 has the supracaudal apparently com- 
pletely divided below, all have the normal pseudo-suture of the supra- 
caudal as seen from above. 

Measurements. The largest, a female, measures 174 mm. in length, 
118 mm. in breadth, and 38 mm. in depth. 

Notes. On December 13 an adult and four young were taken under 
shelter of rocks but exposed. These were shown to the local natives 
who brought in a dozen more, of which nine juveniles were sent alive 
to the Zoological Society of London for experimental feeding with a 
view to developing the exoskeleton; unfortunately only two survived 
the voyage. 



CHELOXIIDAE 

Eretmochelys imbricata (Linnaeus) 
Tesiudo imbricata Linnaeus, 1766, Syst. Nat., Ed. 12, 1, p. 350; American Seas: 
1 (M. C. Z. 30011) Mombasa, Kenya Colony. 9. vii. 30. 

Distribution. This turtle was bought in the native fish-market 
where it had been lying unfed for several weeks. Being inedible, 
Hawksbill Turtles are presumably kept to sell as curios to the occa- 
sional visitors who discover the interesting market which is hidden 
away behind the old customs warehouse at Mombasa. Hawksbill 
Turtles have frequently been recorded from localities on the East 
African coast. 



208 bulletin: museum of comparative zoology 

PELOMEDUSIDAE 

Pelusios sinuatus (Smith) 

Sternothaerus sinuatus A. Smith, 1838, Illus. Zool. S. Africa, 3, pi. i; South 
Africa, "in rivers to the north of 25° S. latitude." i.e. region of the head- 
waters of the Limpopo River. 

Pelusios sinuatus Loveridge, 1929, U. S. Nat. Mus. Bull. No. 151, p. 15: 
Juja Farm, Kenya Colony; Ujiji, Tanganyika Territory. 

2 (M. C. Z. 30012-3) Ujiji, Lake Tanganyika. 28. v. 30. 

Distribution. Sternfeld has recorded this terrapin from Lake Tan- 
ganyika and Roux a juvenile example from Bukoba, Lake Victoria. 

Native name. Fuhce (Kijiji). 

Variation. The remarks made about Raven's specimens from Ujiji 
in 1929 apply with equal force to these examples. I am still uncon- 
vinced that Hewitt's P. s. zuluensis is an\'thing more than a local 
variant but our material is insufficient to form a considered opinion. 
Certainly our three Ujiji terrapin have conspicuous median protuber- 
ances on vertebrals III and IV {i.e. zuluensis) but their two hindmost 
marginals are directed downwards very strongly (as in sinuatus); on 
the other hand a terrapin of the same size from Mt. Chirinda, S. 
Rhodesia recently received from Dr. J. H. Sandground has the pro- 
tuberance on the fourth vertebral even more pronounced than in the 
Ujiji examples (i.e. zuluensis) audits two hindmost marginals are dis- 
tinctly upturned {i.e. zuluensis). I doubt if much reliance can be 
placed on this marginal character as in 250 Testudo tornieri from 
Dodoma every variation was seen, some marginals being so upturned 
as to form a perfect gutter. 

Meas7ircments. Both are young being from 80 to 88 mm. in length. 

Breeding. As Mr. H. C. Raven had secured two terrapin at Ujiji in 
1920 I visited the place in the hope of securing a good series. The 
fishermen, however, did not produce any and told me that Jul}' is the 
month in which they take large numbers of these reptiles. In July, 
they said, the weather and water are warm and the turtles come out 
to lay their eggs. 

Enemies. While some of the Ujiji natives scornfully denied eating 
terrapin, others admitted that they did eat them. 

Habitat. The two specimens collected were found at the bottom of 
a twenty-foot-deep, cement-lined pit in an old ruin. The pit held 
about eighty gallons of stagnant water and this we had to bail out in 
order to secure the reptiles. 



loveridge: African herpetology 209 

Pelusios nigricans nigricans (Dondorff) 

Testudo nigricans Dondorff, 1798, Zool. Betyr. des Linn, natur., 3, p. 34: Type 

locality unknown. 
Sternotherus nigricans nigricans Siebenrock, 1909, Zool. Jahrb. Syst., 3, p. 558. 
Pelusios nigricans nigricans Hewitt, 1931, Ann. Natal Mus., 6, p. 460. 

2 (M. C. Z. 30016-7) Ukerewe Id., Lake Victoria. 11. v. 30. 
2 (M. C. Z. 30018-9) Entebbe, Lake Victoria. 28. vi. 30. 

Variaiion. Heretofore I have followed Siebenrock in referring East 
African terrapin to P. nigricans castaneus {Emys castanea Schweigger, 
1814, Prodr. Chelon., p. 45: Type locality unknown) for he gives the 
ranges as follows : 

Southeast Africa, Rhodesia, Mozambique, Matabele- 

land, East Madagascar n. nigricans 

South Africa, from Natal to the Equator, and from 
the East Coast to the Congo watershed in the West; 

Pemba Island; West Madagascar n. castaneus 

It will be seen that there is a considerable area of overlapping and 
when one attempts to identify material by the s^Tiopsis furnished by 
Siebenrock the results are equally bewildering. For example: 
Anterior border of the second vertebral longer than 

the posterior n. nigricans 

Anterior border of the second vertebral shorter than 

the posterior n. castaneus 

In one Ukerewe specimen (No. 30016) they are equal, in No. 30017 
from the same locality the anterior is a trifle longer; in No. 30018 
from Entebbe they are equal, in No. 30019, also from Entebbe the 
anterior is a trifle shorter. Similar difficulties arise when attempting 
to apply the other characters alleged to differentiate the forms. 
IVIore recently Hewitt has produced a key as follows : 
Shell short, broad and depressed ; marginals V-VII 
with dorsal and ventral surfaces gradually merg- 
ing; intergular shield pear-shaped and longer than 
the anterior border of the humeral ; outer border 

of femoral strongly arched. East Africa P. n. nigricans 

Shell much compressed laterally; marginals V-VII 
having no lateral edge and no definite dorsal and 
ventral surfaces; length of outer border of pecto- 
ral shield a trifle less than, or subequal to, that of 
the outer border of the humeral; intergular shield 
pear-shaped, decidedly longer than the inner 



210 bulletin: museum of comparative zoology 

border of the humeral; outer border of femoral 
slightly arched, and measured in a straight line it 
considerably exceeds the abdominal. Locality? . . P. n. castaneus 

East African terrapin certainly have broad and depressed shells; I 
fail to see any difference in the marginal characters; the shape and 
length of the intergular shield varies greatly; it is, however, pear- 
shaped on the average and in all the score of specimens at my dis- 
posal it is longer than the inner border of the humeral; generally 
speaking, the outer border of the femoral is strongly arched except in 
larger terrapin and occasional specimens where it is certainly only 
slightly arched. I strongly suspect that Boulenger was right in rele- 
gating castaneus to the synonymy of nigricans. 

While terrapin of this genus from a gi^■en locality or lake often ex- 
hibit a family likeness, adequate series from any one place usually 
reveal the fact that few characters are absolutely constant. For 
example Hewitt states that the outer borders of the pectoral and 
humeral shields are equal in nigricans, subequal in P. n. rhodesiamis. 
In a series of six nigricans from Dodoma only three have these outer 
borders equal, in the other tliree the outer border of the humeral is 
much longer than that of the pectoral; in the whole series there is 
much variation in this character. 

Measuremcnis. The greatest shell length is to be found in the 
Ukerewe Island terrapin of which the larger is 235 mm., the smaller 
specimen from Entebbe is 180 mm. 

Enemies. The two examples from Entebbe are only deviscerated 
shells found upon the lake shore; according to native reports these 
terrapin are killed and eaten by the African Sea Eagle {Cuncuma 
vocifer wcifer) a gpecies that is much in evidence on the lake. 



Pelusios nigricans rhodesianus Hewitt 

Pelusios nigricans rhodesianus Hewitt, 1917, Rec. Albany Mus., 3, p. 375, pi. 
21, figs. 2 and 3: Mpika District, Northeast Rhodesia. 

1 (M. C. Z. 30014) Mwaya, Lake Nyasa. 1-8. iii. 30. 
1 (M. C. Z. 30015) Nyamkolo, N. Rhodesia. 9. v. 30. 

Native name. Kajamha (Kinyakusa). 

Affinities. P. n. rhodesianus unfortunately escaped the Zoological 
Record owing to the name being proposed in the bod}' of the text 
under the heading of the typical form only. It was based on a series 
of which the exact number is not stated, from Mpika District which 



loveridge: African herpetology 211 

is about 225 miles due south from Nyamkolo where one of the speci- 
mens listed above was obtained. 

The new form is said to differ from P. nigricans from the Congo as 
figured by Schmidt (1919, Bull. Am. Mus. Nat. Hist., 39, p. 411, 
fig. 1) and also from a Pemba Island specimen figured by Siebenrock 
(1906, in Voekzkow, Reise in Ostafrika, 3, p. 36, pi. 5, fig. 18) under 
the name of S. n. castaneus in the character of the intergular shield 
which is long and narrow in Schmidt's figure, somewhat pear-shaped 
in Siebenrock's, and diamond shaped in rhodesianus of which there 
is an excellently reproduced photograph, and also a figure of a young 
terrapin from Mpika District. 

The elongated, almost parallel-sided, shape of the intergular in 
Schmidt's figured specimen, which was only 67 mm. in length, is 
quite characteristic of the young of P. n. nigricans; with growth the 
sides bulge out until the scale attains the somewhat pyriform shape 
shown in Siebenrock's plate. When our series of nigricans (consisting 
of a score of East African, as well as a Pemba Island and a Western 
Malagasy terrapin) is arranged according to size (46 to 325 mm.) this 
development is readily seen, but none approximate to the appearance 
of rhodesianus more closely than the two examples listed above. The 
two forms may be distinguished as follows: 

Intergular shield not, or but slightly, narrowed 
anteriorly showing a broad free edge on the 
border of the plastron P.n. nigricans 

Intergular shield much narrowed anteriorly being 
nearly excluded from the border of the plastron 
by the gulars P. n. rhodesianus 

In all other respects these rhodesianus agree with our series of 
nigricans, though in the description where "costal viii" is said to be 
"apparently not meeting the abdominal shield in full-grown speci- 
mens," it is obvious that marginal viii was intended. 

Measureinenis. Length of Mwaya terrapin 125 mm. Length of 
Nyamkolo terrapin 110 mm. 

Pelomedusa galeata (Schoepff) 

Testudo galeata Schoepff, 1792, Hist. Testud., p. 12, pi. 3, fig. 1: "Habitat in 
India orientale, Carolina." 

14 (M. C. Z. 30020-33) Mangasini, Usandawi. 13. xii. 29. 

Distribution. A large Helmeted Terrapin was brought in by a 
native when I was camped on Ukerewe Island, 19. vi. 30. Unfortu- 



212 bulletin: museum of comparative zoology 

nately it escaped the same night. I examined a small specimen taken 
on the island which was in the collection of Pere Conrads. The 
species has been recorded from Ukerewe by Sternfeld. 

Native name. Malwala (Chigogo). 

Measurevienis. The largest specimen from Mangasini is a female of 
179 mm., the largest male is 164 mm., the yomigest terrapin is only 
44 mm. 

Ilahiis. The initial downpour of the lesser rains occured at Man- 
gasini on December 12, 1929, and lasting from 6 p.m. till noon on the 
13th brought these terrapin in great numbers from their aestivating 
quarters. Some were encountered in the water course which held a 
torrent the night before; others were taken in freshly-formed pools. 

TYPHLOPIDAE 

Typhlops graueri Sternfeld 

Typhlops graueri Sternfeld, 1912, Wiss. Ergebn. Deutsch-Zentral-Afrika- 
Exped., 4, p. 264: Rain forest behind Randbergen, Belgian Ruanda- 
Urundi. 

2 (M. C. Z. 30034-5) Ujiji, Lake Tanganyika. 30. v. 30. 

Distribution. Hitherto only known from the holotype. The present 
specimens were taken at Ruanda a few miles east of Ujiji and not to 
be confused with Ruanda in the Belgian Mandated Territory a 
hundred and fifty miles to the north. 

Native name. Kisambive (Kijiji). 

Were it not that Ujiji is also the type locality of Amphishaena 
phylofiniens it w^ould be a matter of surprise to find a rain-forest 
species in such a hot and low-lying spot as Ujiji but the process of 
deforestation and desiccation that has been in progress for centuries 
may be actually observed two hundred and fifty miles south of Ujiji 
at Kitungulu, the type locality of Ti/phlojjs gracilis Sternfeld. 

Affinities. T. gracilis only differs from T. graueri in two particulars 
and perhaps the relationship would be better expressed by making 
graueri a race of gracilis. The rostral shields of both are strikingly 
different from that of T. uluguruensis; those of the former, or at least 
of graueri, are enormous and truncated posteriorly, that of uluguruen- 
sis is only "very large" and sharply pointed posteriorly. 

In this connection I should like to draw attention to an unfortunate 
error of transposition in the table of these and allied forms in Barbour 
& Loveridge, 1928, Mem. Mus. Comp. Zool., 50, p. 104 where 1st 



loveridge: African herpetology 



213 



and 2n(l labials under (/radius and iihigurucnsis are transposed; 
correctly restated these should read : — 

Character. T. graueri T. gracilis T. uluguruensis 

Scale rows 24 22 20 

Times the body diameter is 

contained in length 51-61 80 48-51 

The nasal is divided from 

rostral to the 1st labial 1st labial 2nd labial 

Labials with which the ocular 

is in contact 2nd and 3rd 2nd and 3rd 3rd and 4th 

The subocular is absent absent absent 

The snout has a sharp hori- sharp hori- bluntly rounded 

zontal edge zontal edge outline 

The description of graueri being very meagre, the following, drawn 
up independently of it and based on the two I jiji snakes, is given. 

Descriplion. Snout prominent, Avith sharp horizontal edge, nostrils 
inferior; rostral enormous, 3 mm. in length, nearly as long as the 
breadth of the head, extending backwards far beyond an imaginary 
line connecting the commissures of the mouth; eye indistinguishable; 
nostril divided, the suture extending from the 1st labial through the 
nostril to the rostral; no preocular or subocular; ocular small, in con- 
tact with the 2nd and 3rd upper labials; three upper labials; three 
lower labials. Diameter of body contained 51-61 times in the total 
length (60 times in Sternfeld's type); 24 midbody scale rows; tail 
slightly longer than broad in the male, much shorter than broad in 
the female, sharply pointed, the tip spine-like. 

Coloralion. Uniformly flesh-pink in life; colorless or plumbeus in 

alcohol. 

& 9 

Measurements. Length of head and body 300 mm. 202 mm. 

Length of tail 5 mm. 2 mm. 

Diameter at midbody 5 mm. 4 mm. 

The total length of Sternfeld's type was 355 mm. 
Habitat Captured alive by Salimu in an area of low-lying rice 
plantations at Ruanda. 



Typhlops punctatus punctatus (Leach) 

Acontias punctatus Leach, 1819, in Bowdich, Miss. Ashantee, p. 493: Fantee, 
Gold Coast. 

Typhlops tornieri Sternfeld, 1910, Mitt. Zool. Mus. Berlin, 5, p. 69: Kiliman- 
jaro, Tanganyika Territory. 



214 bulletin: museum of comparative zoology 

2 (M. C. Z. 30036-7) Ujiji, Lake Tanganyika. 28. v. 30. 
1 (M. C. Z. 30038) Mwanza, Lake Victoria. 6. vi. 30. 

1 (M. C. Z. 30039) Ukerewe Id., Lake Victoria. 10. vi. 30. 

2 (M. C. Z. 30040-1) Mabira Forest, Uganda. 1. vii. 30. 
1 (M. C. Z. 30042) Jinja, Uganda. 2. vii. 30. 

Distrihution. Recorded from Bukoba by Roux. 

Native name. N dumiakihrili (Kijiji). 

Variation. The eves are indistinguishable in one of the ^Nlabira 
Forest snakes which is obviously about to slough; they present various 
stages of distinctness in the rest of the series; presumably being most 
distinct in those snakes which have sloughed most recently. As a key 
character, the distinctness or otherwise of the eye should be used with 
caution. T. tornieri Sternfeld seems to have been separated solely on 
this character. IVIr. K. P. Schmidt has recently reexamined one of 
the types of tornieri and informs me that he could distinguish the eye 
and that the preocular is realh' in contact with the second and third 
labials. 

In one Ujiji snake (Xo. 30037) the nasal is completely divided on 
both sides ; in the Jinja reptile it is completely divided on the left side 
only, the right side being in the normal condition of incompletely 
divided. This is another key character which cannot be wholly relied 
on. All agree in possessing 28 midbody scale-rows. 

Coloration. The coloration of this series is remarkably uniform and 
answers to Boulenger's variety B. b. (i.e. T. Uneolatus) except Xo. 
30037 where the yellow spots being absent dorsally the entire upper 
surface is dark brown while beneath the yellow spots have coalesced 
to form large blotches resulting in a mottled ventral surface. 

Measurements. The diameter of the body is included in the length 
from 29 to 33 times (24 to 30 in Boulenger, 1893, Cat. Snakes in Brit. 
Mus., 1, p. 42). In size the series ranges from 300 mm. (Mabira) to 
470 mm. (Ujiji). 

Parasites. X'ematodes (Kalicephalus sp.) were found in the snake 
from Ukerewe Island. 

Habitat. The Jinja reptile was taken just before sunset as it was 
wriggling along in a furrow, which served as a gutter, at the side of 
the road. 



Typhlops schlegelii schlegelii Bianconi 

Typhlops schlegelii Bianconi, 1850, Spec. Zool. Mosamb., p. 13, pi. iii, fig. 1 : 
Inhambane, Mozambique. 



loveridge: African herpetology 215 

Onychocephalus dinga Peters, 1854, Monatsb. Akad. Wiss. Berlin, p. 620: Tete; 

Sena; Chupanga, Mozambique. 
Onychocephalus mucruso Peters (part), 1854, Monatsb. Akad. Wiss. Berlin, 

p. 621: Tete, Mozambique. 
Onychocephalus varhts Peters, 1860, Monatsb. Akad. Wiss. Berlin, p. 82: Sena, 

Mozambique. 
(Onychocephalus) riparius Peters, 1881, Sitzber. Ges. naturf Freunde, Berlin, 

p. 50: Chupanga, Mozambique. 

In an attempt to ascertain what name should be applied to certain 
blind snakes collected northeast of Lake X^asa, I found it necessary 
to examine very thoroughly the extremely' involved status of many 
species described from this region. 

The oldest name available was scJilegelii with 40 midbody scale 
rows, the next dinga with 34 to 40. The latter had been referred to 
the synonymy of the former by Bocage but this conclusion was not 
accepted by Boulenger. As no East African material from north of 
the Zambesi possesses more than 38 scale rows and averages much less, 
it became apparent that one might recognize a form in extreme south- 
east Africa, south of the Zambesi, this nominate form being character- 
ized by 34 to 40 (increased to 44 by Boulenger) scale rows. 

In the absence of natural barriers I somewhat arbitrarily refer 
Southern Rhodesian and Transvaal specimens to this race though at 
the present time no examples from these colonies are knoA\Ti to me 
which have more than 38 scale rows, yet they undoubtedly average a 
higher count than do snakes from further north. 

For study 18 snakes of this form were available, 16, including a 
cotype of dinga, in the collection of the Museum of Comparative 
Zoology, and 2 borrowed from the United States National ■Museum 
and the British ^Museum. Of these 3 come from Mozambique south 
of the Zambesi (Tete; Mezi (?Muase) River; Chifumbazi); 13 from 
Southern Rhodesia (Bulawayo; Chikore; Mount Chirinda; Eldorado 
and Kafue River); 2 from the Transvaal (Barberton). 

These IS snakes have from 32-38 midbody scale rows (average 
34.4) ; the rostral is broader than long in 16 snakes, as broad as long 
in 1 snake, longer than broad in 1 snake; the eye is beneath the ocular 
in 5 snakes or beneath the suture between the ocular and preocular 
in 13 snakes; the nasals are separated behind the rostral in 14 snakes, 
or in contact in 4 snakes. The lengths range from IQA to 740 mm.; 
midbody diameters from 6 to 28 mm., the latter being contained in 
the total length from 25 to 35 times. In this connection it might be 
added that the cotype of dinga which, according to Peter's measure- 



216 bulletin: museum of comparative zoology 

ments, had its diameter 41 times in the length, according to my find- 
ings is only 35 times for its diameter is 7 mm., not 6 as stated by 
Peters. 

Ti'PHLOPS scHLEGELii MUCRUSO (Peters) 

Onychocephalus mucruso Peters (part), 1854, Monatsb. Akad. Wiss. Berlin, 

p. 621: Macanga (i.e. Makanga), Mozambique. 
(nychocephalus) petersii Bocage, 1873, Jorn. Sci. Lisboa, 4, p. 249: Biballa, 

Angola. 
Typhlops (Onychocephalus) humbo Bocage, 1886, Jorn. Sci. Lisboa, 11, p. 171: 

Quisange, Benguella, Angola. 
Typhlops mucruso Boulenger (part), 1893, Cat. Snakes Brit. Mus., 1, p. 36: 

Zanzibar; East Africa; Angola. 
Typhlops hottentotus Bocage, 1893, Jorn. Sci. Lisboa, 3, p. 117: Quindumbo, 

Angola. 
Typhlops mandensis Stejneger, 1893, Proc. U. S. Nat. Mus., 16, p. 725: Wange, 

mainland opposite Manda Island, Kenya Colony. Loveridge (part), 

1929, U. S. Nat. Mus. Bull. No. 151, p. 16. 
Typhlops schlegelii Stejneger (nee. Bianconi), 1893, Proc. U. S. Nat. Mus., 16> 

p. 725: Wange, opposite Manda Island, Kenya Colony. 
Typhlops dinga Sternfeld (nee. Peters), 1911, Sitzb. Ges. naturf Freunde, 

Berlin, p. 248: Tabora, Tanganyika Territory. 
Typhlops latirostris Sternfeld, 1910, Mitt. Zool. Mus. Berlin, 5, p. 70: Tabora, 

Tanganyika Territory. 
Typhlops punctatus Loveridge (nee. Leach), 1923, Proc. Zool. Soc. London, 

p. 872, and Barbour & Loveridge, 1928, Mem. Mus. Comp. Zool., 50, 

p. 106: Both records of juveniles from Dar es Salaam, Tanganyika 

Territory.. 

6 (M. C. Z. 30043-8) Bagamoyo. 11-16. xi. 29. 
1 (M. C. Z. 30049) Mangasini, Usandawi. 14. xii. 29. 
11 (M. C. Z. 30050-60) Mwaya, Lake Nyasa. 1-8. iii. 30. 
1 (M. C. Z. 30001) Tukuyu, Rungwe district. 13. iii. 30. 

Native name. Dumilahosa (Kinyakusa). 

Distrihtdio7i. I am now restricting the nominate form, T. s. schleg- 
elii, together with its synonym dinga to East Africa south of the 
Zambesi {vide antea). Ihe next available name is 7H?/cn/50 if we restrict 
its use to the type from Makanga, for mucruso as described by Peters 
was a composite of the two races. This policy is followed as mucruso 
has long been in use for East African specimens and petersii, which 
was next proposed, was long ago referred to the synonymy of mucruso 
by Boulenger. 



loveridge: African herpetology 217 

Syiiotiymy. T. kumho has had rather a complicated history. It was 
recognized as distinct by Boulenger at the time of the publication of 
the Catalogue of Snakes, but having no topotypic specimens he 
incorrectly assumed that a Mpwapwa snake which he had was 
identical and substituted his own description of the Mpwapwa snake 
for that of Bocage's Angolan reptile. I have seen two topotypes of 
humho which is undoubtedly a synonym of miicniso, a conclusion 
reached by Boulenger in 1915. I have also seen the Mpwapwa snake 
which I refer to T. schlegelii excentricus Procter. 

T. mandensis. This snake was described as from Wange, Manda 
Island and collected by Gustave Denhardt. Denhardt's plantation 
at Wange is, however, not on the island but on the mainland twenty 
miles north of Manda Island and there are no topographical grounds 
therefore for considering it different from mucruso which occurs along 
the coast further south. 

From 1893 to 1923 no second specimen was reported until I recorded 
one from Morogoro, which I now consider was a T. schlccjcUi exccn- 
tricu^s that was blind and colorless because about to slough. This 
Morogoro snake is in the British Museum. 

The characters in which inandensis and mucruso were supposed to 
differ may be contrasted as follows: — 

inandensis mucruso 

(a) Horizontal edge of snout obtusely angular sharply angular 

(b) Eyes hidden distinct 

(c) Diameter of body into length 23 times 23 to 38 times 

(d) Midbody scale-rows 34 30-38 

(a) As shown below under " pimctatus,'" an obtusely angular snout 
is normal in the young of mucruso, but becomes sharply angular in the 
adult. The type of mcmdensis, which I have examined, is a young 
animal measuring 135 mm. 

(b) Elsewhere (1923, Proc. Zool. Soc. London,' p. 873) I have drawn 
attention to a T. s. excentricus which, when taken at Kilosa, was 
"whitish, or flesh-colored, w^ith the eyes completely hidden but after a 
period of captivity was found to be normally colored and the eyes 
distinct." It therefore seems probable that prior to sloughing the old 
epidermis becomes opaquely white and the eyes invisible. The color 
description of mandcnsis lends support to this view, "Uniform pale 
greenish-gray above, pale buff beneath." The Morogoro snake re- 
ferred by Boulenger and myself to " mandensis" was still more so for 
it was " colorless except for a little buff on the belly." 



218 bulletin: museum of comparative zoology 

(c) At the time of the publication of the description of mandensis 
the range of diameter into length for mucruso was only 25 to 35 times, 
it has since been extended and includes that of mandensis. 

(d) The midbody scale-rows have always been within the recog- 
nized range of variation for mucruso, even in 1893. 

T. " schlegelii." The snake referred to schlegelii by Stejneger was 
collected with the type of mandensis at AYange. It is half-growTi and 
has 34 scale-rows. I consider it identical with m.ucruso which has 
been reported from Lamu (near Wange) by Sternfeld. I have also 
examined the snake from the Lado Enclave which Boulenger referred 
to schlegelii, it has 38 scale-rows which is high for so northerly a speci- 
men, in other respects it is identical with Uganda mucruso. 

T. lotirostris. Sternfeld described this snake on the basis of an 
individual which was colorless and about to slough. He referred other 
Tabora specimens to mucruso and dinga. Boulenger synonymised 
latirostris with mucruso in 1915. 

T. " punctatus." Possibly Sternf eld's record of this species from 
Tabora is based on a juvenile as were mine from Dar es Salaam. A 
few years ago the Museum of Comparative Zoology received from Dr. 
J. H. Sandground a series (M. C. Z. 29167-29174) of eight snakes from 
Mt. Chirinda, Southern Rhodesia, which range in size from 195 to 
740 mm. I refer these to T. s. schleglii. I had never seen so fine a 
developmental series from one locality and was immediately struck 
by the similarity of the smallest to the three young Dar es Salaam 
snakes which I had previously referred to punctatus, the intermediate 
to half-grown snakes such as those listed above, and the largest to 
mucruso as figured by Peters in,"Reise nach Mossambique," 1882, 
plate xiii, fig. 3. (The locality of the figiu-ed specimen not being stated 
it cannot be said whether it is mucruso as here restricted, or one of the 
Tete or Sena specimens referred to T. s. schlegelii, but this is im- 
material to the discussion.) 

I found that the snout in the smallest Chirinda snakes was obtusely 
angular as in adult jmnctatus but became progressively sharply angu- 
lar with advancing age till it presents the appearance figured in Peter's 
mucruso. A comparison of the small Dar es Salaam snakes (130 and 
296 mm.) with young Chirinda specimens (195 and 275 mm.) showed 
them specifically identical in every respect, even to details of colora- 
tion. They are regarded as racially distinct as specimens from south 
of the Zambesi do attain a higher scale-count. 

Variation. The 19 snakes listed above have from 30-36 midbody 
scale-rows (average 32.1); the rostral broader than long in the adults, 



loveridge: African herpetology 219 

about as broad as long in the young; the eye, when visible, is beneath 
the ocular or beneath the suture between the ocular and preocular; 
the nasals are separated behind the rostral in 17 snakes, or in contact 
in 2 snakes. 

Seven snakes were borrowed from the British Museum for com- 
parison. These came from Zanzibar; Mombasa, Kenya Colony; 
Uganda; Lado Enclave; Lake Tanganyika and Quisange, Angola and 
had been previously identified variously as T. schlegelii, T. mucruso, 
T. varius and T. kumbo. 

These 7 snakes have from 30 (Mombasa) to 38 (Lado and Quisange) 
midbody scale-rows (average 35.1); the rostral is broader than long 
in all 7 snakes; the eye is beneath the ocular; the nasals narrowly or 
broadly separated behind the rostral. The lengths range from 273 to 
610 mm.; midbody diameters from 11 to 23 mm.; the latter being 
contained in the total length from 23 to 34 times. 

It may be possible to differentiate a form with very large rounded 
rostral like a trench helmet as opposed to a moderately large rostral 
with lateral sides almost parallel, so many intermediate conditions 
occur that I failed, nor did this variation appear to occur with any 
geographical significance. In Peter's figures are shown two types of 
head, one in which the head passes gradually into the body, the other 
in which the head seems disproportionately small. These snakes lay 
up stores of fat, presumably for aestivation through the long dry 
season, and I suggest that this is responsible for the swollen bodies of 
some specimens; I may be WTong. 

Measuremenis. Twenty-two Tangan^'ika snakes measure from 132 
to 485 mm. in total length with midbody diameters of from 5 to 15 
mm., being contained in the total lengths from 26 to 38 times as 
against 25 to 37 given by Boulenger in his 1915 key. 

Diet. Ants were preserved from the stomach of one Bagamoyo 
snake while two leathery snake's eggs, measuring 14 x 6 mm. were 
found in the stomach of another. 

Defence. An adult male emitted a very strong-smelling caecal 
discharge when first captured. 

Habitat. This adult was actually taken at Kitopeni about five 
miles south of Bagamoyo where it was dug from sand at the base of 
a banana plant; a half-growTi snake was taken in sand under debris 
beside a young coconut palm. It would appear as if very large adults, 
which are rarely encountered, live deeper underground than the 
smaller snakes, only coming to the surface when the first rains fall 
after a long dry season. 



220 bulletin: museum of comparative zoology 



Typhlops schlegelii excentricus Procter 

Typhlops humbo Boulenger (nee. Bocage) part, 1893, Cat. Snakes Brit. Mus., 

1, p. 46: Mpwapwa, Ugogo, Tanganyika Territory. 
Typhlops excentricus Procter, 1922, Ann. Mag. Nat. Hist. (9), 9, p. 685: 

Kilosa, Tanganyika Territory. Loveridge, 1923, Proc. Zool. See. London, 

p. 874: Kilosa. 
Typhlops mandensis Loveridge {nee. Stejneger), 1923, Proc. Zool. Soc. London, 

p. 872: Morogoro, Tanganyika Territory. 
Typhlops dinga Loveridge {nee. Peters), 1923, Proc. Zool. Soc. London, p. 873 

and 1929, Bull. Antivenin. Inst. Amer., 3, p. 14: Kilosa and Morogoro, 

Tanganyika Territory. 
Typhlops mucruso var. humbo Loveridge {nee. Bocage), 1923, Proc. Zool. 

Soc. London, p. 873: Ilonga; Kidenge; Kilosa; Kipera; Madazini and 

Mpwapwa, Tanganyika Territory. 
Typhlops mucruso Barbour & Loveridge {nee. Peters), 1928, Mem. Mus. 

Comp. Zool., 50, p. 109: Kilosa; and Loveridge (part), 1929, U. S. Nat. 

Mus. Bull. No. 151, p. 17: Morogoro specimen only. 

1 (M. C. Z. 30062) Morogoro, Ukami. 30. xi. 29. 

Distribution. This color form is recognized because it is confined to 
a definite area in Central Tanganyika Territory in which Typhlops 
sMcciclii mucruso is not known. 

J'ariation. As a result of the knowledge gained as to age variation 
in T. s. schlegelii, I gathered together all the available blind snakes 
which I had collected at Morogoro, Kilosa and Mpwapwa and at 
villages in their vicinity. These consisted of thirteen snakes of which 
seven were topotypes of excentricus, in addition the data from the 
description of excentricus was added as well as the data from the 
British Museum specimen from Mpwapwa which was referred to 
humbo by Boulenger in 1893. 

These fourteen snakes have from 30-36 midbody scale-rows; the 
rostral is broader than long below, or as broad as long; the eye, when 
visible, is beneath the ocular in 7 snakes or beneath the suture be- 
tween the ocular and preocular in 7 snakes; the nasals are just in 
contact behind the rostral in 10 snakes or separated behind the rostral 
in 4 snakes. 

It will be seen, therefore, that there is no scale character by which 
one may separate excentricus from mucruso and hence from 1923 
onwards they have been considered s;v'nonymous. The large series of 
mu-cruso now available, however, leads me to separate the two forms 
on a basis of their coloration. This action had already been taken by 



loveridge: African herpetology 221 

Sternfeld in 1910 though, following Boulenger, he erroneously applied 
the name humho to e.r centric us, which had not been described at that 
time. 

In T. s. excentricus the lower surface is colored and spotted like 
the back, such a form of coloration is only knowTi to me from Moro- 
goro, Kilosa, Mpwapwa and vicinity. T. s. mucruso on the other 
hand usually has the lower surface entirely white, or yellow, or at 
least a longitudinal median area of white, the upper surface being 
extremely variable. 

In describing T. excentricus, the late Miss Procter compared it with 
mucruso believing that it could be separated by the nasals which were 
in contact behind the rostral and by the shape of the rostral which 
was truncated posteriorly instead of rounded. Both conditions occur 
in snakes from Kilosa and Morogoro and are not specifically important. 
The type had 30 midbody scale-rows and the diameter was included 
in the length 44 times. 

In 1923, by closely following the key to the genus Typhlops in 
Boulenger's "List of the Snakes of East Africa" (1915, Proc. Zool. 
Soc. London, p. 614) I was able to split these Kilosa and Morogoro 
snakes into four " species " which demonstrates that the key is based 
on characters which are not specifically differentiating. 

The reasons for referring a IMorogoro snake to T. mandensis will be 
obvious from a perusal of my reasons for considering T. mandensis 
Stejneger a synon\Tn of T. s. mucruso. The Morogoro snake, pre- 
sumably being about to slough, showed no traces of an eye and in 
consequence falls into the WTong section of the key. Boulenger con- 
firmed this identification and later, when I UTOte to him from East 
Africa, he very kindly reexamined the specimen and replied that he 
still considered it answered to the description of mandensis. 

The name dinga was applied to two large snakes whose diameter 
into body length agreed with the large type of that species and whose 
dorsal coloration they closely resembled. 

Measurements. These fourteen snakes measure from 135 to 860 
mm. in total length with midbody diameters of from 5 to 20 mm., 
being contained in the total lengths from 22 to 47 times. 

Summary of races. 

Midbody scale-rows 32 to 42 (rarely 44) ; diameter 
included in length from 25 to 45 times; at least 
median line of belly immaculate. (East Africa 
south of the Zambesi) T. s. schlegelii 

Midbody scale-rows 30 to 36 (rarely 38) ; diameter 



222 bulletin: museum of comparative zoology 

included in length from 23 to 38 times*; at least 
median line of belly immaculate. (Angola, Cen- 
tral and East Africa north of the Zambesi except 

for a small area in Central Tanganyika) T.s. mucruso 

Midbody scale-rows 30 to 36 ; diameter included in 
length from 22 to 47 times; belly colored like the 
back. (Mpwapwa to Morogoro in central Tan- 
ganyika Territory) T.s. excentricus 

Typhlops opisthopachys Werner 

Typhlops opisthopachys Werner, 1917, Mitt. Zool. Mus. Hamburg, 34, p. 35: 
Tanga, Tanganyika Territory. 

On first reading the description of this snake I observed that it 
was totally unrelated to any group of African Typhlops. After an 
exhaustive attempt to find any Ethiopian allies, I came to the con- 
clusion that it was an Australian species. 

I wrote to my friend Dr. Werner and asked whether he could 
furnish any further information as to its origin. He replied that the 
snake had been received from "a German engineer who indicated 
that he had found it at Tanga." It is now in the Hamburg Museum. 
I suggest that possibly the donor was an engineer on a German liner 
who may have forgotten in what port he obtained the snake in ques- 
tion. 

My reasons for making such a suggestion are as follows. This 
Typhlops is a peculiar species of unusual proportions and yet Werner's 
description agrees in almost every detail with a specimen of Typhlops 
pinguis Waite (1897, Trans. Roy. Soc. South Austral., 21, p. 25, pi. 
iii: South Australia) from Lake Preston, West Australia (M. C. Z. 
32813). 

The only points in which Werner's description differs from the 
Australian reptile are as follows: 

Ocular above the 3rd and 4th supralabials ; rostral 
almost as broad as the nasal, at its broadest about 
half the breadth of the head, at its narrowest in- 
cluded three and a half times in the breadth of the 

head pinguis 

Ocular above the 3rd and 4th supraocular; rostral as 
broad as the nasal, included four times in the 
breadth of the head opisthopachys 

♦Undoubtedly will be extended when iMger specimens are available. 



loveridge: African herpetology 223 

As the ocular cannot be above its supraoculars, the latter is evidently 
a misprint for supralabials. The other differences I propose to dis- 
regard in view of the complete agreement in all other respects of these 
two peculiarly proportioned snakes. 

Therefore I propose to consider opisthopachys Werner a synonym of 
pinguis Waite and remove it from the African list as has already been 
done with Fanchonia elegans Werner (gen. et sp. nov.) which was 
found to be based on the Australian Hyla aurea (Lesson). 

LEPTOTYPHLOPIDAE 

Leptotyphlops emini (Boulenger) 

Glaucoma emini Boulenger, 1890, Ann. Mag. Nat. Hist. (6), 6, p. 91: Karagwe, 
Bukoba, Tanganyika Territory. 

1 (M. C. Z. 30063) Kitungulu, Urungu. 15. v. 30. 

Distribution. This species has been recorded by Boulenger from 
Nyamkolo (Niomkolo) to the south of Kitungulu. 

Native 7iame. Luminuviinu (Kirungu). 

Measurements. Total length 102 (94.5+7.5) mm., the diameter, 
which is 2 mm., is included in the length 51 times. 

Habitat. Taken by digging in sandy soil beneath a log on a hillside. 

Leptotyphlops longicauda (Peters) 

Sienostoma longicauda Peters, 1854, Monatsb. Akad. Wiss. Berlin, p. 621, 

Tete, Mozambique. 
Glauconia emini Loveridge (nee. Boulenger), 1923, Proc. Zool. Soc. London, 

p. 874: Dar es Salaam, Tanganyika Territory. 
Glauconia longicauda Loveridge, 1923, Proc. Zool. Soc. London, p. 875: Lumbo, 

Mozambique; Angel, 1925, in Voyage de Ch. Alluaud et R. Jeannel en 

Afrique Orientale (1911-1912), p. 30: Kulumuzi, Tanga, Tanganyika 

Territory. 

The unfortunate misidentification of a young longicauda as emini 
was due to placing too much reliance on the generic key in Boulenger's 
"List of the Snakes of East Africa." (1915, Proc. Zool. Soc. London, 
p. 616). The appearance of Angel's first record of the occurrence of 
longicauda in Tanganyika caused me to examine his snake and re- 
examine my Dar es Salaam specimen. As a result I suggest the follow- 
ing amendment to the key which should read : — 

Diameter of body 45-57 times into total length; tail 

8-11 times; color black emini 



224 bulletin: museum of comparative zoology 

Diameter of body 57-70 times into total length; tail 

8-9 times; color white (pink in life) longicauda 

Leptotyphlops conjuncta (Jan) 

Stenostoma conjuncta Jan, 1861, Arch. Zool. Anat. Fisiol., 1, p. 189: South 

Africa. 
Glaucoma conjuncta Boulenger, 1893, Cat. Snakes Brit. Mus., 1, p. 67: South 

Africa; Natal; Kilimanjaro, Tanganyika Territory. 
Glaucoma scutifrons Lonnberg (nee. Peters), 1907, Reptilia and Batrachia in 

Sjostedt, 1910, Kilimandjaro-Meru Expedition, 1, part 4, p. 14: Ngare na 

nyuki, Tanganyika Territory. 
Glaucoma merkeri Werner, 1909, Jahresh. Ver. Nat. Wurttemb., 65, p. 61: 

Moshi, Tanganyika Territory; Loveridge, 1923, Proc. Zool. Soc. London 

p. 874: Mtali's village, Mkalama, Tanganyika Territory. 
Glaucoma latirostris Sternfeld, 1912, Wiss. Ergebn. Deutsch-Zentral-Afrika- 

Exped., 4, p. 264: Northwest of Lake Tanganyika. 
Glaucoma emini Sternfeld {nee. Boulenger), 1912, Wiss. Ergebn. Deutsch- 

Zentral-Afrika-Exped., 4, p. 264: Ukerewe Island, Tanganyika Territory. 
Glauconia distanti Loveridge {nee. Boulenger), 1923, Proc. Zool. Soc. I^ondon, 

p. 874: Morogoro and Kilosa, Tanganyika Territory. 
Leptotyphlops distanti Barbour & Loveridge {nee. Boulenger), 1928, Mem. 

Mus. Comp. Zool., 50, p. 109; Loveridge, 1929, U. S. Nat. Mus. Bull. 

No. 151, p. 18: Morogoro and Mt. Longido, Tanganyika Territory. 

1 (M. C. Z. 30064) Entebbe, Lake Victoria. 28. vi. 30. 

2 (M. C. Z. 30065-6) Ukerewe Id., Lake Victoria. 10. vi. 30. 

Affinities. The key referred to under the last species is still more 
unfortunate in making a major division based on whether the rostral 
shield extends backwards beyond the level of the eyes for it may or 
may not do so in conjuncta though it always does so in distanti ( = scuti- 
frons); the width of the rostral in relation to that of the head is rather 
difficult to define in practice. This key misled Boulenger himself in 
referring my first ]VIorogoro snakes to distanti, a course which I have 
consistently followed with all East African material where the rostral 
did so extend. In the larger L'kerewe snake it extends further than in 
the smaller and this caused me to reopen the whole question; while 
passing through London I took the opportunity of comparing the 
larger Ukerewe snake with the type of distanti and found that though 
the rostral extends backwards in both snakes in a similar degree yet 
tlie Ukerewe snake had not nearly so ln*oad a rostral as the type of 
distanti. East African records of the occurrence of distanti or scutifrons 
north of the Rovuma River should be referred to conjuncta. 



LOVERIDGE: AFRICAN HERPETOLOGY 225 

AYhen, in 1912, Sternfeld reported L. cmini from Ukerewe Island 
he stated that his specimen was so dried up that he could not be certain 
of the identification. .\s my specimens from that island are un- 
doubtedly coniuncta I amend his determination. 

In describing G. latirostris in the same paper he states that it only 
differs from conjuncta in possessing a larger rostral extending beyond 
the level of the eyes, a character which the Ukerewe snakes show to 
be variable with age. 

Through the exceeding kindness of Dr. Wilhelm Gotz of the Wurt- 
temburg Xaturaliensammlung, Stuttgart, I have been able to examine 
the two cotypes of Werner's Glauconia mcrkcri from Moshi at the foot 
of Kilimanjaro. Having carefully measured and remeasured these 
specimens several times I find that Werner was in error in his measure- 
ments. Actually the 160 mm. snake even when stretched is only 150 
mm., and the 175 mm. specimen only 172 mm., the diameter of both 
is approximately the same, 2.75 mm., resulting in the diameter being 
included in the total length only 54-62 times instead of 80-87 times 
as reported; the tails are included in the length 10.7 to 13.2 times. Nor 
can I agree that the rostral is " at least twice as broad as the nasal," 
I should say that it is not quite twice as broad. The posterior edge of 
the rostral is about level, or slightly beyond, an imaginary line con- 
necting the posterior borders of the eyes but as I have stated above 
this condition is common in L. conjuncta of which I consider L. merkeri 
a synonym. 

Having reached these conclusions I commvmicated with ]\Ir. V. 
FitzSimons regarding the status of distanti of which they possess a 
good series of topotypic (Pretoria, Transvaal) material in the Trans- 
vaal ]\Iuseum. Under date of January 29, 1931 he replied, " . . .1 
have found extreme difficulty in separating scutifrons and distanti. I 
have gone very carefully over our series and cannot find any distinc- 
tive characters on which to separate them. The average proportions 
of diameter of body and of length of tail, into total length, work out 
about the same, while the rostral varies so much in shape and size that 
it cannot be used as a satisfactory character. I have, however, one 
specimen in which the rostral covers practically the whole head, but 
owing to its bad state of preservation little else is distinguishable. 
This may be distanti, but until I can obtain further material I am re- 
garding distanti as a synonym of scutifrons." 

The series of scuiifrons in the collection of the Museum of Compara- 
tive Zoology were likewise inseparable from a pair of snakes from 
Pretoria received as distanti so that I believe we are justified in con- 



226 bulletin: museum of comparative zoology 

sidering disfanfi a s\Tionym of scutifrons which was described from 
Sena, Mozambique. 

My final conclusion is that conjunda and scutifrons may be most 
readily separated by the width of the rostral in its relation to the 
nasal, thus : — 

Rostral at least three times the width of the nasal ; diam- 
eter of body 42 to 68 times into the total length scutifrons 

Rostral not more than twice the width of the nasal; diam- 
eter of body 32 to 72* times into the total length conjuncta 

Measurements. The larger Ukerewe snake, a male, measures 142 
(130+12) mm., the smaller 74 mm. Diameters are 2 and 1.5 mm. and 
are included in the total length 71 and 49 times respectively. 

Enemies. The Entebbe snake was recovered from the stomach of a 
young burrowing viper (Atractasjns irregularis) and the head being 
digested away the identification is based on an Entebbe specimen of 
conjuncta in the British Museum. 



BOIDAE 

Python sebae (Gmelin) 

Coluber sebae Gmelin, 1788, Syst. Nat., 1, p. 1118: No type locality. 
Python sebae Boulenger, 1893, Cat. Snakes Brit. Mus., 1, p. 86. 

Skulls and skin (M. C. Z. 30067-9) Ukerewe Id., Lake Victoria. 16. vi. 30. 

Measurements. With a view to obtaining data which might prove of 
assistance in estimating the actual length of a snake whose dried skin 
only is available, I measured one of these snakes in the flesh and 
found it to be 2,180 mm., while its dried, and not unduly stretched, 
skin measured no less than 2,650 mm. That is to say an increase of 
at least .21 of the total length should be allowed for, or in other words 
a dried skin is nearly a quarter as long again as was the living reptile 
from which it was taken. The skull of this same snake measures 84 
mm. in its greatest length so that it may be assumed that a python is 
about twenty-six times longer than its skull though this proportion 
varies with age for the larger snake measured 4,330 mm. in the flesh 
with a skull length of only 128 mm. or a thirty-third of its total length. 

Diet, folklore etc. Ukerewe Island is somewhat famous for its big 
pythons; the large dimensions which they reach may be attributed to 
the beliefs of the Wakerewe who object to the killing of these snakes 

*:52 times in a bloated female from Wakkerstroom, Transvaal, though other snakes from the 
same locality are from 40-60 times. 72 times in a snake from Mt. Lougido. 



loveridge: African herpetology 227 

for they hold that death or misfortune will befall the slayer or his 
relatives as a consequence of his action. 

A few months prior to my arrival on Ukerewe, Mr. W. E. H. Scup- 
ham, District Officer of Mwanza, had visited the island and shot a 
python there. He had planned to return to Mwanza the same evening 
but the engine of his motor launch broke down and, having neither 
oars nor sail, he and his men drifted about all night till the currents 
brought them back to Ukerewe Island in the morning. "There you 
are," said Chief Gabriel, "that comes of your killing this python." A 
few weeks later a further communication from Gabriel reported that 
a python had caught and killed a woman on the island. I was asked 
by Scupham to investigate this report. Curiously enough on the very 
evening of my arrival at Murutunguru, Pere Conrads of the Catholic 
Mission of Marienhof, himself a well-known naturalist, showed me 
the head of this snake and communicated to me the details of how it 
had killed the woman. 

The woman had been engaged in washing clothes beside a stream 
and spreading them out upon the ground nearby. She was not very 
well at the time having, only eight days previously, given birth to a 
baby which had died. A native coming to the ford observed the 
clothes spread about but no sign of the owner; he called but received 
no reply. Thinking this strange he began a search in the vicinity and 
came upon the woman lying dead in the coils of a huge python. Re- 
turning to the village he summoned the men who, overcoming their 
usual reluctance, killed the reptile with four spear thrusts and two 
knifings. The snake was measured and found to be four and a half 
metres, wdth a midbody diameter of forty centimetres. 

The natives stated that thirty years ago a youth or big boy {kijana 
kuhwa) was killed by a python on the island. This is the only fatal 
case in the recollection of the old men who said that though many 
persons have been caught by pythons they invariably escape by exert- 
ing their strength. An educated Mkerewe told me that so great was 
the aversion to killing a python that should one of his fellow tribesmen 
find his own child dead in the coils of one of these snakes he would not 
kill the snake for he would argue, "The child is dead anyway, why 
should I die also for killing the reptile." Tangible evidence of their 
dread of dead pythons was observed when the specimen listed above 
was being skinned by the side of the mainroad; several natives were 
seen to retrace their steps on catching sight of it afar off and then they 
made an extensive detour rather than pass within a hundred feet of 
the remains. 



228 bulletin: museum of comparative zoology 

On Thursday evening, June 12, 1930, the village headman told me 
that two natives had seen a big python out in the bush. On my en- 
quiring why they had not come and reported the fact to me at once, 
he replied that they feared I w^ould shoot and skin it, in which case 
some calamity might befall them. I told him to inform them that I 
promised not to shoot it, nor even to hit it but would take it alive. 
Next day I was returning to camp for lunch at 4 p.m. when I was told 
that the natives had been back to the place where they had seen the 
python and found it still in the same spot; they had come to report 
this to me but had grown tired of waiting as I was so long absent! I 
sent for them but only one appeared ; first of all he asked if it was true 
that I promised not to kill it and on being assured that this was the 
case he next inquired what I would give him for revealing the where- 
abouts of the snake, which I was assured was quite close. Forgetting 
for the moment that there were two men to be compensated I replied, 
"Fifty cents." (being equivalent to 12c in U. S. currency or 6d in 
English coinage). He murmured something and I continued my meal. 
When I had finished and enquired for the man, it was said that he had 
gone off saying that fifty cents was insufficient, he would catch the 
snake himself, perhaps, and sell it to me. I replied that I was quite 
willing to do this but if he hit it with a stick I should refuse to buy it 
as I wanted it alive and well. The tiresome business was eventually 
concluded by the man's return and my fresh offer of fifty cents to 
each of them. This was entirely satisfactory; the missing companion 
was immediately forthcoming and we set out at 5 p.m. armed with a 
T-shaped stick, a small sack eighteen inches square and, as I had 
nothing else available, my duffle bag in which to put the python when 
captured. 

Instead of being "nearby," as stated, we had to W'alk a mile and a 
half through the bush before we reached the place, and when we got 
there the python was gone! It had been lying in a dense thicket over- 
growTi with rank grass taller than a man, the spot where it had been 
was plain enough for the grass was flattened where it had lain. The 
thicket covered an area of about thirty feet by twenty feet and there 
were many similar thickets in the immediate vicinity. 

Pushing into the thicket I found that the snake had only moved 
some six feet further in for I caught sight of its coils which were 
enormous, at midbody the girth was greater than that of an average 
man. I proceeded to beat down the grass and brambles till one had a 
clear view and while I was so engaged the reptile struck at me open- 
jawed, his head going as high as my chest and once as high as my 



loveridge: afbican herpetology 229 

throat. Having cleared the arena, I hung the small sack on the end 
of the snake stick and let the snake strike the sacking half-a-dozen 
times; each time it struck, the sack would fall to the ground only to 
be lifted again on the end of the stick. While I was engaged in recover- 
ing it the snake struck too quickly a couple of times and to avoid the 
blow I had to step back so that the force of it was spent in space; the 
snake flopped down but quickly withdrew its head on the defensive 
again. After a dozen futile attempts in one of which it struck the end 
of my snake stick through the sack — and the blow had the force of a 
sledge hammer — the python became discouraged and decided to 
retire. As it commenced to glide away I sprang after it, planted 
the T-end of my stick on its neck for a second as I grasped it firmly 
with my hand before it had time to throw ofT the stick. Salimu, who 
had been standing by waiting for orders, now came running to seize 
its tail; Abedi grasped it round the middle and a temporary employee 
held open my duffle bag crying continuously in the vernacular, "It 
cannot go, it won't go in." Thrusting the snake's head to the bottom 
of the bag I seized the neck again from the outside, i.e. through the 
material. Before going in, however, it dribbled from both ends, try- 
ing, I think, to disgorge the bushbuck which it had swallowed, the 
result was the most appalling stench imaginable. It would have made 
most people sick but at the moment we had other things to think 
about! We just crammed and crammed that poor old python into the 
bag and after we succeeded in doing this it was all we could do to lace 
up the opening with some cord which I had brought for the purpose. 
By this time it was nearly dusk and as I had much to attend to in 
camp, I gave orders to cut two poles (and the boys had only one jack- 
knife among them with which to accomplish this) and so convey the 
duffle-bag to camp. As soon as the poles were ready the two guides 
refused to carry the snake on superstitious grounds. I had left sup- 
posing that they would lend a hand. It was impossible for my three 
boys to carry the two poles, so they did what seemed to them the next 
best thing and slung the bag from a single pole. Unfortunately they 
wound the bark-cord round and round the middle of the bag at one 
spot. In this way they carried it for about a mile until both Abedi's 
shoulders were raw, then they gave up and sent one of their number 
on to camp for help. When they at last reached camp with their load 
it was after dark, and we immediately placed the bag on the mission 
scales, Pere Conrads supervised the weighing and found the pj'thon 
scaled sixty kilos (135 lbs.). Then I emptied the bag on to a comfort- 
able bed of straw in a large packing case, nailed down the lid, put a 



230 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 

fifty kilo weight on top of the lid, and left it for the night about 
twenty feet from my camp bed but where I could see it in the moon- 
light from where I lay. 

During the night a Spotted Hyena approached very near but scared 
by my moving, decamped and did not return, it had evidently got a 
whiff of my captive. Purposely I left the snake for a few days to digest 
its meal intending to pack it carefully on the third day as I was leaving 
Ukerewe Island on the fourth. When the third day arrived and I 
opened the box it was to find the python dead and already so far gone 
in decomposition that only the skull was worth saving. We stretched 
it out, however, and found that it measured fourteen feet, four inches 
(4,330 mm.) in length, while the diameter at midbody was one and a 
half feet (440 mm.), the actual circumference being three feet, three 
and a half inches (1,000 mm.), nor was this astonishing size due to 
inflation resulting from decomposition for dissection revealed a fully 
adult bushbuck doe in milk, if not in young. The reason for the 
python's death was obvious, the rhythmic jerk-jerk to which the bag 
had been subjected as the boys staggered along with it had caused the 
snake to be strangulated by the cord v,-ound round the middle of the 
bag which threw all the weight on one spot. 

At noon on June 18th an Mkerewe came to tell me that a python 
which had taken one of his goats some weeks before had just been 
located near his village. It was two miles to the place; when we got 
there, after a trying walk in the midday heat and glare, the snake was 
found lying more or less extended, part of its length being concealed 
in a bush. It was captured without the least difficulty; its length was 
approximately ten feet yet it weighed thirty-eight kilos when boxed, 
probably thirty kilos or sixty-five pounds net. It was boxed at Mwanza 
and dispatched on June 23rd by rail to Dar es Salaam. On July 10th 
I opened the box on board the S.S. Usamhara and found that the 
snake had made no evacuation but it did so as soon as I lifted it from 
the box. I transferred it to a tub of fresh water, covered the tub with 
a sack, and left it to soak for an hour before returning it to its box. A 
week later I again removed it for an hour's soaking and found that it 
had sloughed and was resplendent with a beautiful bloom on its scales. 
The box was shallow, about a foot in depth, and was lined with sacking 
which was padded on sides and bottom with excelsior, the lid was a 
series of slats but the interspaces protected by sacking. In this re- 
ceptacle the snake travelled to London arriving in excellent condition, 
being presented to the London Zoological Society (which had de- 
frayed all expenses) by the Museum of Comparative Zoology. 



loveridge: African herpetology 231 

COLUBRIDAE 

Xatrix olivaceus (Peters) 

Coronella olivacea Peters, 1854, Monatsb. Akad. Wiss. Berlin, p. 622: Tete, 
Mozambique. 

3 (M. C. Z. 30070-2) Mwaya, Lake Nyasa. 1-8. iii. 30. 
1 (M. C. Z. 30073) Nkuka Forest, Rungwe Mtn. iii. 30. 
1 (M. C. Z. 30074) Albertville, Lake Tanganyika. 21. v. 30. 
1 (M. C. Z. 30075) Ukerewe Id., Lake Victoria. 19. vi. 30. 

Xatirc name. Injalalu (Kinyakusa). 

I 'ariation. The mountain snake, as might be expected, had 17-17-L5 
scale-rows, the Albertville, unexpectedly, 17-17-17, the Ukerewe 
Island reptile 17-19-17 while the three ]Mwaya specimens were 
normal in possessing 19-19-17, the foremost row in these might be 
21 if counted almost on the back of the head. 

Ventrals ranged from 131-142; anals divided; subcaudals 53-63 
but only those of three snakes were countable the others having lost 
the ends of their tails; the subcaudal range of forty Uluguru and 
Usambara snakes was 63-87. 

Coloration. The Rungwe Mountain snake agreed with Uluguru and 
Usambara specimens in having the edges of the ventrals plum-colored 
or reddish-mauve while the central area was bright orange. Both the 
Albertville and Ukerewe snakes possessed a vertebral stripe of choco- 
late brown, agreeing in this respect with the series from Stanleyville, 
Belgian Congo in the collection of the American ^Museum of Natural 
History but differing from all other East African olivaceus which 1 
have seen. 

Measuremenis. All are of small size, the largest snake being 460 
(360+100) mm., the smallest 206 (161+45) mm.; both are from 
Mwaya. 

Diet. The stomach contents of a ]Mwaya snake consisted of a frog 
{A rthroleptis mi nut us) . 

Enemies. Three of the series have lost their tails, this is 50% of the 
total therefore rather higher than in mountain snakes. 

Habitat. One day, being caught in a heavy downpour while on the 
flats near INIwaya, we sought refuge in a small hut at the edge of a 
rice field. The hut, which was used only when the rice was ripening, 
was in a very delapidated state and its floor strewn with grass. More 
than a score of natives gathered in this hut for shelter when suddenly 
an uproar arose — a native had found himself sitting upon a couple of 



232 bulletin: museum of comparative zoology 

snakes! One of these escaped me, the other I secured and it proved to 
be an Olive Water Snake. The Albertville snake was taken under a 
bundle of thatching grass lying beside a half-finished hut which was 
also situated in a swampy flat. 

GLYtHOLYCus BicoLOR Giinther 

Glypholycus bicolor Giinther, 1893, Proc. Zool. Soc. London, p. 629: Lake 

Tanganyika. 
Herophidion hypsirhinoides Werner, 1924, Sitzb. Ak. Wiss. Wien, 133, p. 53: 

"? New Guinea." 

1 (M. C. Z. 30076) Sumbwa, Lake Tanganyika. 20. v. 30. 

Variation. Agrees in e\'ery detail with the description of the type 
series; it has been compared with a topotype received in exchange 
from the British Museum. 

Measurements. A male, measuring 481 (365+110) nim. 

Habitat. This specimen was found freshly-dead, washed up on a 
sandbar of the Lake shore. Local natives stated that Glypholycus, 
like Boulengerina, only occurs on the rock-bound coasts but that there 
was such a rocky foreshore a few miles from Sumbwa where Glypholy- 
cus was plentiful. 

BoAEDON LiNEATUS Dumeril & Bibron 
Boaedon lineatus Dumeril & Bibron, 1854, Erpet. Gen., 7, p. 363: Gold Coast. 



1 
1 
1 
1 
11 
1 
3 
1 
1 
1 
1 
1 
1 
1 



M. C. Z. 30078) Bagamoyo. 16. xi. 29. 

M. C. Z. 30079) Unyanganyi, Turn. 4. xii. 29. 

M. C. Z. 30080) Dabaga, Uzungwe Mtns. 1. i. 30. 

M. C. Z. 30077) Madehani, Ukinga Mtns. 14. ii. 30. 

M. C. Z. 30081-91) Mwaya, Lake Nyasa. 1-8. iii. 30. 

M. C. Z. 30092) Tukuyu, Rungwe. 13. iii. 30. 

M. C. Z. 30093-5) Ilolo, Rungwe. 31. iii. 30. 

M. C. Z. 30096) Igale, Poroto Mtns. 30. iv. 30. 

M. C. Z. 30097) Nyamkolo, Lake Tanganyika. 9. v. 30. 

M. C. Z. 30098) Ujiji, Lake Tanganyika. 28. v. 30. 

M. C. Z. 30099) Mwanza, Lake Victoria. 6. vi. 30. 

M. C. Z. 30100) Ukerewe Id., Lake Victoria. 19. vi. 30. 

M. C. Z. 30101) Entebbe, Uganda. 27. vi. 30. 

M. C. Z. 30102-3) Kampala, Uganda, vi. 30. 



Distribution. The Brown House Snake has been recorded from 
Mpwapwa and Bukoba, as well as Uhehe and Ukerewe Island, by 
Sternfeld. 



loveridge: African herpetology 233 

Native names. Melawuletzi (Kikinga) ; injoka (Kinyakusa, but not 
specific). 

Variation. ^Nlidbody scale-rows 27-31; ventrals 194-231 (males 
194-207, females 210-231); anal single; subcaudals 41-67 (males 
61-67, females 41-53); labials 8, the 4th and 5th entering the orbit 
excepting for Nos. 30077-8 where it is the 3rd, 4th and 5th; preoculars 
2 except for Nos. 30077, 30079 and 30093 where there is only a single 
preocular so that Parker (1930, Ann. Mag. Nat. Hist. (10), 6, p. 598) 
is incorrect in stating that a single preocular is normal for this species, 
at least so far as East African snakes are concerned; the upper pre- 
ocular is in contact with the frontal in thirteen specimens, on one side 
of the head only in two, and separated from the frontal in eleven. 

Measurevients. The largest snake measures 681 (560+121) mm. 
and is from Mwaya; the biggest female is 970 (860+110) mm. and was 
taken at Ilolo; the smallest snake, also from Mwaya, measures 255 
(221+34) mm. 

Sex. The proportion of males to females is ten to fifteen. 

Diet. Stomach contents consisted of rodents as follows: (1) Arvican- 
this abyssinicus vmanzae at Mwanza, (2) Cryptoimjs hoUentotvs whytei 
at Ilolo, and other unidentified rodents in Ilolo, Mwaya and Madehani 
snakes. 

Parasites. A tick was removed from the throat of the Nyamkolo 
snake; nematode worms (Kalicephahis sp.) were found in its stomach 
and also in the stomach of an Unyanganyi reptile which had in addi- 
tion an undescribed species of Arduenna. 

Enemies. One Brown House Snake, found dead on the road, pre- 
sumably killed by a car, was put into a cage containing three young 
Banded Mungoose (Mungos mungo colonus) which immediately 
attacked and devoured all of it except the head and backbone. 

Lycophidion capense capense (Smith) 

Lycodon capense A. Smith, 1831, S. Africa Quart. Journ., (1) No. 5, p. 18: 
Kurrichane, i.e. Rustenberg district, Transvaal. 

4 (M. C. Z. 30109-12) Mwaya, Lake Nyasa. 1-8. iii. 30. 
1 (M. C. Z. 30113) Ujiji, Lake Tanganyika. 28. v. 30. 
1 (M. C. Z. 30114) Ukerewe Id., Lake Victoria. 10. vi. 30. 
1 (M. C. Z. 30115) Kampala, Uganda, vi. 30. 
1 (M. C. Z. 30116) Jinja, Uganda. 30. vi. 30. 

Distribution. The Cape Wolf Snake has been recorded by Sternfeld 
as occurring at Bagamoyo, Dar es Salaam and Ukerewe Island ; while 



234 bulletin: museum of comparative zoology 

he lists another from Entebl)e but under the name of L.jacksoui which 
is a synonym. Roux reports it from Bukoba. 

Xativc name. Micani/alulosha (Kinyakusa). 

Variation. Midbody scale-rows 17; ventrals 174-211 (extremes 
from Kampala and Mwaya); subcaudals 29-48 (Jinja and Mwaya); 
labials 8, the 3rd, 4th and 5th entering the orbit with the exception of 
the Ukerewe Island snake which has only 7 labials of which the 3rd 
and 4th enter the orbit, Gth largest. Some might wish to refer it to 
the West AiTicaji fasciahim but it is undoubtedly an aberrant cajpense. 

Coloration. All the above agree in having the throat more or less 
white, which is also the case with twenty-two other specimens in the 
Museum of Comparative Zoology from Kenya, the Usambara and 
Uluguru Mountains, Rhodesia, the Transvaal, Cape Colony and 
Southwest Africa, this in contrast with L. c. acutirostre. 

Measurements. The largest male measures 507 (352 -|- 155) mm., 
and the largest female 497 (445-|-52) mm., both from Mwaya. 

Lycophidion intermediates between capense and acutirostre 

G tint her 

? Lycophidium acutirostre Boettger in Voeltzkow, 1913, Reise in Ostafrika, 3, 
pt. 4, p. 363: Mavene, near Tanga, Tanganyika Territory. Sternfeld, 1912, 
Wiss. Ergebn. Deutsch-Zentral-Afrika Exped., 4, p. 268: Kenya Colony. 

5 (M. C. Z. 30104-8) Bagamoyo. 11-12. xi. 29. 
also the following material : — 

1 (M. C. Z. 5992) Zanzibar. 

1 (M. C. Z. 18191) Kilosa, Tanganyika Territory. 

1 (M. C. Z. 18192) Morogoro, Tanganyika Territory. 

Variation. Midbody scale-rows 17; ventrals 158-169; subcaudals 
27-42; if the Bagamoyo series be taken alone the ventrals range is 
158-169 and the subcaudals 27-33. 

With the exception of a single snake taken by Sir John Kirk shortly 
after he secured the type series of four, no additional specimens of 
L. acutirostre Giinther have been recorded from that island so far as I 
am aware. Sternfeld has considered a snake from Kenya Colony 
with 159 ventrals and 33 subcaudals referable to this species. The 
five Zanzibar snakes, however, ranged from 140-150 in number of 
ventrals and 18-28 in subcaudals, a sixth snake from Zanzibar with 
179 ventrals and 45 subcaudals was referred to L. capense by Boulenger 
in 1893. 



loveridge: African herpetology 235 

In 1915 Boulenger records the recognized scale-counts as 
L. capeme 163-208 ventrals and 24-47 subcaudals 
L. acutirostre 140-150 " " 18-28 

from which it will be seen that the present series are intermediate 
though in the dark coloring of their throats they agree with acutirostre. 

It seems probable, therefore, that acutirostre represents the extreme 
of variation in the extreme east (Zanzibar) of the extensive range of 
capense and that occasional Zanzibar snakes, as well as those from 
Bagamoyo, Morogoro and Kilosa must be regarded as intermediates. 
It might be added that the Morogoro-Kilosa fauna is essentially that 
of the coastal belt. 

Coloration. All the above agree in having the throat and lower 
surface uniformly blackish-brown. 

Measurements. The largest Bagamoyo snake measures 344 (320+ 
24) mm. 

Diet. The following skinks were recovered from three Bagamoyo 
snakes: — (1) Riopa s. sioidcvallii, (2) Ablepharus wahlhergii, (3) four 
tails of A. wahlhergii, the latter is of interest as showing the preserva- 
tion value of a readily fractured tail! 

Parasites. Both nematodes and tapeworms {Oochoristica sp.) were 
recovered from these snakes. 

Lycophidion capense uzungwensis Loveridge 

Lycophidion capense uzungwensis Loveridge, 1932, Bull. Mus. Comp. Zool. 
72, p. 375: Dabaga, Uzungwe Mtns., Tanganyika Territory. 
c? (M. C. Z. 30117) Dabaga, Uzungwe Mtns. 1. i. 30. 
9 (M. C. Z. 30118) Kigogo, Uzungwe Mtns. 23. i. 30. 

Distribution. As the localities from which these snakes came are in 
the far north and extreme south of the Uzungwe range this race should 
occur throughout the mountains. I heard of a third specimen from a 
German lady who, on learning that I was collecting snakes, asked if I 
"had ever seen a black snake with a bright red arrow-like marking 
on its head?" She added that she had seen one in these mountains 
and was struck by its unusual appearance. 

Diet. The female held a skink {Mabuya varia varia) in her stomach. 

Pseudaspis cana (Linnaeus) 

Coluber canus Linnaeus, 1758, Syst. Nat., ed. 10, 1, p. 221: "Indiis." 

1 (M. C. Z. 30119) Lukungu, Ubena Mtns. 8. ii. 30. 
1 (M. C. Z. 30120) Mangoto, Ukinga Mtas. 10. ii. 30. 



236 bulletin: museum of comparative zoology 

Native name. Ketumba (Kikinga). 

Variation. Both normal in possessing 25 midbody scale-rows; 192- 
200 ventrals; divided anals and 56-46 subcaudals. 

Measurements. Both are young, the Lukungu snake, presumably a 
male with fewer ventrals and more caudals, measures 297 (250+47) 
mm., and the Mangoto female 290 (255+35) mm. 

Habitat. I captured the Lukungu snake in a marshy valley at an 
altitude of about 6,000 feet. 

Chlorophis emini (Giinther) 

Ahaetulla emini Gunther, 1S88, Ann. Mag. Nat. Hist. (6), 1, p. 325: Monbuttu, 
Belgian Congo. 

1 (M. C. Z. 30121) Mwaya, Lake Nyasa. 1-8. iii. 30. 

Distribution. This record considerably extends the known range for 
the most southerly known to me are those of Roux and Sternfeld for 
Bukoba. 

Native name. Imbindipindi (Kinj'akusa for green tree snakes). 

Variation. Midbody scale-rows 15; ventrals 158; subcaudals 103; 
9 labials, the 4th, 5th and 6th entering the eye. It only differs from 
the type in having 103 instead of 111 subcaudals and in possessing 
two preoculars. 

Measurements. Total length 800 (550+250) mm. 

Chlorophis hoplogaster (Gunther) 

Ahaetulla hoplogaster Gunther, 1863, Ann. Mag. Nat. Hist. (3), 11, p. 284: 
Port Natal, i.e. Durban, South Africa. 

1 (M. C. Z. 30122) Bagamoyo. 18. xi. 29. 

3 (M. C. Z. 30123-5) Mwaya, Lake Nyasa. 1-8. iii. 30. 

5 (M. C. Z. 30126-30) llolo, Rungwe. 15-30. iii. 30. 

1 (M. C. Z. 30131) Ukerewe Id., Lake Victoria. 16. vi. 30. 

Distribution. Sternfeld has recorded Jwplogaster from Kitopeni; 
Bagamoyo; Tukuyu and Bukoba, and Boulenger from Victoria 
Nyanza. 

Native name. Imbindipindi (Kinyakusa for green tree snakes). 

Variation. Midbody scale-rows 15; ventrals 141-160; subcaudals 
82-95; labials 8, the 4th and 5th entering the eye except on the right 
side of No. 30129 where the 4th, 5th and 6th enter. The IIolo snakes 
with from 141-150 ventrals are lower than the others which are 150 



loveridge: African herpetology 237 

for all three Mwaya snakes, 155 for Bagamoyo and 160 for Ukerewe 
Island; the range for the species as recognized hitherto was 150-169. 
Number 30125 and the Ilolo series are intermediate between hoidogas- 
ter and neglectus for the former have definite, even if slight, traces of 
ventral keels while of the latter three are quite smooth, one has scat- 
tered traces of keeling on the ventrals and two more pronounced 
keeling. I have noted a similar intermediate condition in a Kenya 
snake (U.S.N.M. 49012) which leads me to suspect that neglectus is 
more entitled to be regarded as a race of hoplogaster than as a full 
species. 

Coloration. One Ilolo snake had the neck mottled pale blue and 
brown, another, said to have been taken within the Nkuka Forest 
boundaries, possessed paired blue spots on either side of the mid- 
dorsal area on the anterior third of the back as was the case with 
snakes referred to neglectus from the rain forest of the Uluguru Moun- 
tains. 

Measurements. The Ukerewe Island snake measures 712 (500+212) 
mm., all the series from the southwest are small, the largest is only 
550 (380+170) mm. 

Breeding. Five ova, measuring 28 x 8 mm., were taken from a 
Mwaya snake on March 3; the large Ukerewe snake held only two 
eggs measuring 29 x 8 mm. on July 16, 1930. 

Diet. A gecko (Hemidacti/lus persiviilis) was recovered from a 
Bagamoyo snake and a frog (Arthroleptis whytii) from an Ilolo reptile. 

Chlorophis neglectus (Peters) 

Philothamnus neglectus Peters, 1866, Monatsb. Akad. Wiss. Berlin, p. 890: 
Prazo Boror, Mozambique. 

& (M. C. Z. 30132) Kigogo, Uzungwe Mtns. 23. i. 30. 

Distribution. Already recorded from Uhehe by Sternfeld. 
Native name. Nyaluwina (Kihehe). 

Variation. Midbody scale-rows 15; ventrals definitely with lateral 
keels 145; subcaudals 88; labials 8, the 4th and 5th entering the eye. 
Measurements. Total length 513 (360+153) mm. 
Diet. A frog {Arthroleptis reichei) in its stomach. 

Philothamnus semivariegatus semivariegatus Smith 

Philothamnus semivariegatus A. Smith, 1849, Illus. Zool. S. Afr. 3, pis. 59, 60, 
and 64: Bushman's Flats and Kurrichane, S. Africa. 



238 bulletin: museum of comparative zoology 

4 (M. C. Z. 30133-6) Bagamoyo. 14-16. xi. 29. 

1 (M. C. Z. 30137) Mangasini, Usandawi. 14. xii. 29. 

1 (M. C. Z. 30138) Kigogo, Uzungwe Mtns. 30. i. 30. 

1 (M. C. Z. 30157) Kipili, Lake Tanganyika. 19. v. 30. 

2 (M. C. Z. 30158-9) Ukerewe Id., Lake Victoria. 14. vi. 30. 

Distribution. Previously recorded by Sternfeld from the Zanzibar 
coast, Dunda in Kingani, Ukerewe Island etc. 

Native names. Mlaluwe (Kihehe); namafwa (Kifipa). 

Variation. Midbody scale-rows 15; ventrals 172-198; anal divided; 
subcaudals 134-156; labials 9-12 (10 on one side of No. 30157 and 
12 on one side of No. 30159) with the 4th, 5th and 6th entering the eye 
except in the Kigogo snake which has the 5th and 6th only and Nos. 
30157 and 30159 which have the 5th, 6th and 7th on the left side of 
the head only; temporals only uniform on both sides of the head in four 
snakes, treating the sides separately therefore fourteen show 2+2, 
three 2+1 and one only 1+2. 

Measurements. The largest snake is a male (No. 30137) measuring 
1205 (800+405) mm., the biggest female (No. 30138) being 1175 
(790+385) mm.; all are of large size. 

Diet. Bagamoyo snakes held (1) Hemidactijlus mabouia, (2) //. 
persimilis, (3) Indeterminate gecko remains, (4) Ablepharus wahlbergii. 
An Ukerewe snake had swallowed two geckos {Lygodadylus picturahi^ 
var.). 

Philothamnus semivariegatus dorsalis (Bocage) 

Leptophis dorsalis Bocage, 1866, Jorn. Sci. Lisboa, 1, p. 69: Duque de Braganca 

and Molembo, Angola. 
Philothamnus dorsalis Schmidt, 1923, Bull. Am. Mus. Nat. Hist., 49, p. 78: 

Banana, Lower Congo. 

22 (M. C. Z. 30139-56) Mwaya, Lake Nyasa. 1-8. iii. 30. 

Distribution. Hitherto considered Angolan and Lower Congo in 
range, this form has not been recorded previously from Tanganyika 
Territory but the record is not incompatible with the knowTi occur- 
rence of other Angolan vertebrates in the area south of Lake Tan- 
ganyika and east of Lake Nyasa. Schmidt not unnaturally assumes 
that Tornier's record of this species from "Lubwa's Usoga" in "Der 
Kriechthiere Deutsch-Ost-x\frikas" is a Tanganyika record but Fort 
Lubwa is on the east bank of the Nile in Usoga, L^ganda. "Lubwa's" 
presumably means "Chief Lubwa's village." 

Native name. Imbindipindi (Kinyakusa for green snakes). 



loveridge: African herpetology 239 

J'ariaiion. It should be remarked that these snakes agree with 
semivariegatus in coloring for they lack the browTi vertebral line 
possessed by Angolan dorsalis. At the same time their lepidosis agrees 
with dorsalis rather than semivariegatus. 

Midbody scale-rows 15; ventrals 167-179; anal divided; subcaudals 
134-156; labials 8-10 (only 8 in two specimens and 10 in one) usually 
the 4th, 5th and 6th entering the e^e but as the condition is azygous 
in six snakes a summary shows that four sides have the 4th and 5th 
entering, twenty the 4th, 5th and 6th, eighteen the 5th and 6th; one 
snake with a damaged head is omitted from this count. 

P. semivariegatus has been differentiated from P. dorsalis as usually 
having 2+2 temporals though 2+1, 1+2 and even 1 + 1 occur very 
occasionally. P. dorsalis on the other hand has 1 + 1 or rarely 1+2, 
all other scale counts are within the range of P. semivariegatus. This 
Mwa^'a series is rich in intermediates, in fact onlv twelve of the 
twenty-one snakes have a uniform arrangement of temporals on both 
sides of the head. Treating them separately, therefore, it is seen that : 

28 sides have 1 + 1 temporals 
11 " " 1+2 

No. 30147 is the only snake with 2+2 on both sides of its head and 
should, perhaps, be called P. s. semivariegatus. In these circumstances 
it seems advisable to treat dorsalis as a race of the older form. Bou- 
lenger has recorded a single P. semivariegatus from the heart of Angola. 

Measurements. The largest snake (No. 30151) measures 965 (610+ 
355) mm., the smallest 421 (275+146) mm. 

Diet. A tree frog (Megalixalus fortuwinii) was the onl}' identifiable 
food found in the stomachs of these arboreal snakes. 



Hapsidophrys lineata Fischer 

Hapsidophrys lineatus Fischer, 1856, Abh. Natur. Ver. Hamburg, 3, p. Ill, 
pi. 2, fig. 5: Elmine, West Africa, i.e. Elmina, Gold Coast; Sternfeld, 1912, 
Wiss. Ergebn. Deutsch-Zentral-Afrika Exped., 4, p. 270; Kindu, Belgian 
Ruanda. 

Gastropyxis orientalis Werner, 1909, Stuttgart Jahresh Ver. Natk. Wurttemb., 
65, p. 56: German East Africa. 

Hapsidophrys lineata Loveridge, 1929, U. S. Nat. Mus. Bull. 151, p. 24: 
Kisumu, Kenya Colony. 

For some years, I have been hoping for the opportunity of examin- 
ing the type of G. orientalis which Dr. Werner informed me was in the 



240 bulletin: museum of comparative zoology 

Stuttgart Natural History INIuseum but which the Director of that 
Institution tells me cannot be found. The type constituted the only 
example of the genus Gastropyxis found in Tanganyika Territory or 
Kenya Colony and its description coincides so exactly with that of 
H. lineata that I have little compunction in assuming that its author 
inadvertently referred it to the wrong genus. 

Ilapsidophrys and Gastropyxu are distinguished in Boulenger's key 
solely by the absence of keels and notches on the subcaudals in Hajj- 
sidophrys, these being present in Gastropyxis; though this difference 
is plain enough when specimens are contrasted the subcaudals of 
Hapsidophrys are so angular (though not notched) that it is very easy 
to assume that they are keeled, in fact it was through finding some 
Hapsidojjhrys mislabelled Gasirojyyxis m a collection that lead me to 
suspect the affinities of orientalis. I might add that the maxillary 
teeth are more numerous in Hcqmdophrys than in Gastropyxis. 

Coronella semiornata Peters 

Coronella semiornata Peters, 1854, Monatsb. Akad. Wiss. Berlin, p. 622: Tete, 
Mozambique. 

& (M. C. Z. 30160) Mangasini, Usandawi. 16. xii. 29. 

Distrihuiion. Sternfeld has already recorded the Semiornate Smooth 
Snake from Mbugwe to the northeast of Mangasini, and I have listed 
two from Kipetu to the southwest. 

Variation. In all respects normal; midbody scale-rows 21; ventrals 
186; anal divided; tail-tip missing; labials 8, the 4th and 5th entering 
the eye. 

Diet. Its stomach held a skink {Riopa sundevallii modestuvi). 

Grayia tholloni Mocquard 

Grayia tholloni Mocquard, 1897, Bull. See. Philom. Paris, (8) 9, p. 11: French 
Congo; Boulenger, 1909, Proc. Zool. Soc. London, p. 951, fig. 

9 (M. C. Z. 30161) Ukerewe Id., Lake Victoria. 11. vi. 30. 

Distribution. Boulenger has recorded this snake from Entebbe and 
my collectors have previously taken it at Bukoba; the type of G. 
fasciata Boulenger, which the author subsequently relegated to the 
synonymy of tholloni, came from southwest Lake Tanganyika. 

When wading in grass-growTi marshes fringing Lake Tanganyika at 
Nyamkolo, several times I disturbed snakes which darted through 



loveridge: African herpetology 241 

the weeds and water with great speed. I failed to secure any but 
concluded that they were G. tholloni. 

Variation. This specimen agrees with Boulenger's revised descrip- 
tion except that the temporals are 2+2, instead of 2+3, and the lower 
anterior temporal is longer than its distance from the frontal. Midbody 
scale-rows 15; ventrals 141; anal divided; tail-tip missing; labials 8, 
the 4th and 5th entering the eye. 

Breeding. This semiaquatic snake, though 510 mm. in length from 
snout to vent, held only three eggs; these measured 25 x 16 mm., and 
were developing, possibly indicating that the species is viviparous. 

Diet. The stomach held the hind legs of a large frog {Rana ? mas- 
careniensis) . 

Parasites. Worms (Kalicephalus sp. and Thuhunea sp. n.) were also 
recovered from its stomach. 



DUBERRIA LUTRIX SHIRANUM (Boulcnger) 

Homalosoma shiranum Boulenger, 1894, Cat. Snakes Brit. Mus., 2, p. 276, pi. 

13, fig. 1 : Shire Highlands, Nyasaland. 
H. lutrix var. atriventris Sternfeld, 1912, Wiss. Ergebn. Deutsch-Zentral- 

Afrika-Exped., 4, p. 271 : Eassenje, Belgian Ruanda. 

4 (M. C. Z. 30162-5) Dabaga, Uzungwe Mtns. 1. i. 30. 

1 (M. C. Z. 30166) Ihanganya, Uzungwe Mtns. 6. i. 30. 

16 (M. C. Z. 30167-82) Kigogo, Uzungwe Mtns. 13-30. 1. 30. 
4 (M. C. Z. 30183-6) Mangoto, Ubena Mtns. 10. ii. 30. 

2 (M. C. Z. 30187) Tandala, Ukinga Mtns. 11. ii. 30. 

9 (M. C. Z. 30188-96) Madehani, Ukinga Mtns. 14-28 ii. 30. 
6 (M. C. Z. 30197-202) Ilolo, Rungwe. 28. iii. 30. 

Native names. Nyaluhereka (Kikinga.) ; isaJcani (Kinyakusa). Both 
the AVakinga and Banyakusa, however, consider this small snake is 
the }'Oung of Trivierorhinus tritaeniatvs to which the}' apply the same 
names. 

Affinities. H. shiranmn has paragraph priority over //. ahysshiicum 
of the same author who referred both to the synonymy of H. lutrix 
(Linnaeus) a few years later. The type locality of lutrix is given as 
"Indiis," in all probability it came from the Cape of Good Hope. 

As there are only ten examples of typical hdrix in the Museum of 
Comparative Zoology I solicited the aid of Mr. V. FitzSimons of the 
Transvaal IMuseum who very kindly supplied me with the data of 
thirty-six specimens in his care. 

Of forty-six specimens of lutrix in the Museum of Comparative 



242 BULLETIX: MUSEUM OF COMPARATIVE ZOOLOGY 

Zoology and Transvaal Museum, forty -three have a loreal on both 
sides of the head, only two (from Tokai, near Cape Town and St. 
Lowry's Pass) lack this shield on both sides of the head while T. M. 
No. 5167 has no loreal on the right side only; thirty-nine snakes have 
2 postoculars on both sides of the head, two have 2 on the left but 
only 1 on the right while five have 1 postocular on both sides of the 
head (from Port Elizabeth; Natal; and 3 Transvaal localities); of one 
of these Mr. FitzSimons states that it is accompanied by eight young 
all of which are typical lutrix. 

Of the forty-two specimens listed above and one from Arusha, also 
in Tanganyika Territory, forty-one have no loreal, the Arusha and one 
Ilolo snake possess a loreal; thirty-three snakes have 1 postocular, 
five have 2 on one side of the head only (from Ihanganya; Kigogo and 
Ilolo) and five have 2 postoculars on both sides of the head (Kigogo). 
That is to say the great majority of snakes from north of the Zambesi 
may be differentiated from their southern relatives by the absence of 
a loreal and presence of a single subocular. Under these circum- 
stances I feel justified in reviving the name shirfnivm in a subspecific 
sense, thus : — 

A loreal (absent in 7%) and two postoculars (or 
1 only in 16%); a dark vertebral band usually 

present; range south of the Zambesi lutrix lutrix 

No loreal (present in 4%) and only one postocular 
(or 2 present in 18%) ; no definite vertebral band; 

range north of the Zambesi lutrix shiranuvi 

Sternfeld has listed some intermediates from the Central African 
lake region and has proposed the name of atriventris for one with a 
black undersurface, a loreal and a single postocular. The Kigogo 
series shows that there is no significance in the color of the belly which 
is very variable. H. I. atriventris must be regarded as a s^Tionym of 
D. I. shiranuin. 

Variation. Midbody scale-rows 15; ventrals 126-151; anal single; 
subcaudals 24-46; labials 6, the 3rd and 4th entering the eye in 38 
snakes, the 2nd, 3rd and 4th in Nos. 30179 and 30191, the 3rd, 4th 
and 5th in No. 30186, while No. 30188 is normal on the left side but 
has 7 labials with the 3rd, 4th and 5th entering on the right; thirty- 
two snakes have a single postocular, five have a single shield on one 
side and two on the other, five have two on both sides of the head. 
Coloration. Great variation was observed in the Kigogo series and 
this was found to be uncorrelated with sex. The following notes were 
made in the field. 13. i. 30. Male. Abo^•e, olive extending to the 



loveridge: African herpetology 243 

outer edges of the ^•ent^als. Below, uniformly pale yellow. Female. 
Above, black. Below, white or greyish-white with irregular black 
blotches. 23. i. 30. Of four females brought in two were black and 
one a rich red-brown with a vertebral series of minute black stripes. 
30. i. 30. Four females brought in present astonishing variation in 
color. 

Measuremenis. The largest male is from Mangoto and measures 
379 (300+79) mm., the largest female is from Kigogo and measures 
412 (362+50) mm., the smallest snake, also from Mangoto, is only 
118 (93+25) mm. in length and can have been born but recently for 
Barbour and I have recorded an embryo of 100 (85+15) mm. from 
Lulonga, Belgian Congo. 

Sex. One is at once struck by the greater length of the tails in 
the males and this is correlated with a higher number of subcaudals; 
the majority of snakes were sexed in the field and I think that the 
following key is a fairly accurate means for distinguishing sex in this 
region. It may be that some few males have fewer than 40 subcaudals 
(Boulenger has recorded one with 35 from South Africa under Inirix 
and his ahyssinicum type had 32) but in general they range higher. 

Length of tail included in total length 4.2 to 5.1 times, 

caudals usually over 40 (41-46 in above series) males 

Length of tail included in total length 6 to 9 times, caudals 

always less than 40 (25-38 in above series) females 

Breeding. At Kigogo, on 23. i. 30, three females holding eggs, viz. 
(1) 13 eggs measuring 11 x 6.5 mm., (2) 9 eggs measuring 9x6 mm., 
(3) 4 eggs measuring 10 x 6 mm. At the same locality but on 30. i. 30 
four more females were taken holding 12, 11, 9 and 8 eggs respectively, 
all approximated to 12 x 8 mm. and were larger, therefore, than in the 
snakes taken a week before. It was observed that a very good ratio 
was kept between the size of the snake and the number of ova which 
it held, the smaller snakes having proportionately fewer ova developing. 

At Madehani, on 14. ii. 30, five females with 12, 10, 8, 7 and 7 eggs 
respectively, those of one of the latter measuring 10 x 6 mm.; there 
was very little sun at Madehani during the whole month. 

Diet. These small snakes subsist almost exclusively upon slugs, 
even a very young Ilolo reptile disgorged a slug when captured and 
slugs were taken from the stomachs of snakes caught at Kigogo, 
Mangoto and Madehani. In addition to a slug I removed a 125 mm. 
Duherria I. shiranum from the stomach of a Madehani snake. 

Para.sites. Nematode worms were recovered from both the stomachs 
and intestines of snakes taken at Dabaga and Kigogo. 



244 bulletin: museum of comparative zoology 

Enemies. Cannibalistic as related above. 

Habitat. The first specimens captured at Dabaga were found in 
rich short grass in marshland in a broad valley. They were presumably 
basking on the top of the grass tussocks where their olivaceous color 
rendered them inconspicuous; as one approached they \ATiggled into 
the roots of the grass where they were almost impossible to find. 

At Kigogo a large series was secured by nati\'es engaged in hoeing 
over grassland on a hillside in preparation for extensive planting by 
the forest officer. The Ihanganyi snake was taken by myself as it 
was about to cross the path in grassland at sunset. 

Near Kigogo a settler, whose native employees were clearing land 
at the edge of forest for planting coffee, informed me that "blind 
snakes" were very abundant and that in digging a furrow forty feet 
in length they had destroyed over ninety of them! As I found no 
Typhlops in the Uzungwe jVlountains I concluded that he referred to 
Duberria. In view of the great economic value of these snakes in a 
coffee plantation by reason of their diet of slugs they deserve protec- 
tion by all intelligent settlers; the same applies to Typhlops which 
subsists almost entirely on termites (white ants) with an occasional 
caterpillar or slug. 

Duberria I. shiranum fills a niche in the East African fauna which is 
occupied by Sforeria dekayi in the North American. There is con- 
siderable similarity in external appearance and size, in diet, in habitat 
and hiding places, and in their ovoviviparitj'. 



Prosymna ambigua Bocage 

Prosymna ambigua Bocage, 1873, Journ. Sci. Lisboa, 4, p. 218: Duque de 
Braganga, Angola; Barbour & Loveridge, 1928, Mem. Mus. Comp. 
Zool. 50, pp. 121-122: East Africa. 

Prosymna variabilis Werner, 1909, Jahresh. Nat. Ver. Wurttemb., 65, p. 57; 
Moshi, Tanganyika Territory. 

d" (M. C. Z. 30203) Bagamoyo, ii. xi. 29. 

cf (M. C. Z. 30204) Mpwapwa, Ugogo. 22. xi. 29. 

Distribution. Sternfeld has recorded ambigua from Zanzibar and 
Bukoba. 

Affinities. I am still unconvinced that the name ambigua is the 
correct one to apply to East African snakes for all that I have seen 
have had 15 scale-rows while ambigua (Angola) had 17; perhaps the 
name bocagii (Congo) should be revived as a race of ambigua and 



loveridge: African herpetology 245 

applied to Congo and East African snakes with 15 scale-rows. I have 
no western material to enable me to form an opinion of any value. A 
full discussion of the relationships of East African snakes to bocagii 
will be found in the 1928 citation given above. 

Due to the kindness of Dr. Wilhelm Gotz, I have been enabled to 
examine the 9 cotype of Werner's P. variabilis and have no hesitation 
in referring it to the synonymy of amhigua (or rather bocagii). The 
chief claim of P. variabilis to specific rank was its rounded snout 
which is prominent, sharp and slightly upturned in bocagii. Hewitt 
has stated with regard to his series of P. transvaaletisis that in the 
young the rostral is not so angular as in the adults. Obviously the 
rounded snout is ancestral and its shovel-like development is corre- 
lated with subsequent adaptation to burrowing habits. The 9 cotype 
of variabilis is so young that the ventral scutes in the umbilical region 
are still unhealed, it measures 95 (86+9) mm. and the male cotype 
(whose whereabouts I have been unable to trace) was only 122 mm. 
Both agreed with the type of bocagii in all essentials though the antero- 
lateral angles of the frontal do not reach the eyes in variabilis, a 
character which has been sho^vn to be variable. The ventral counts of 
the types of variabilis are 140-143 and that of the type of bocagii 167 
but both are within the recognized range of variation. 

Variaiion. Midbody scale-rows 15; ventrals 136-137; subcaudals 
31-33; labials 6, the 3rd and 4th entering the eye; postoculars 1 except 
on the right side of the head of the Mpwapwa snake where there are 
2, while 2 is normal for ambigua, 1 is by no means uncommon. 

Measurements. The larger snake, from Bagamoyo, measures 205 
(175+30) mm., the smaller one only 130 (112+18) mm. 

Habitat. The ]MpAvapwa snake was taken in sandy debris among the 
rotten roots of a fallen tree lying fifty feet from a stream; owing to 
prolonged drought the ground was dry and dusty at the time. 

DASYPELTINAE 
Dasypeltis scaber (Linnaeus) 
Coluber scaber Linnaeus, 1766, Syst. Nat., 1, p. 384: Indiis. 

1 (M. C. Z. 30205) Maji Malulu, Usandawi. 10. xii. 29. 

1 (M. C. Z. 30206) Mwaya, Lake Nyasa. 1-8. iii. 30. 

2 (M. C. Z. 30207-8) Ukerewe Id., Lake Victoria. 10. vi. 30. 

Distribution: Recorded by Sternfeld from Dar es Salaam, Mpwapwa, 
etc., by Roux from Bukoba. 



246 bulletin: museum of comparative zoology 

Variation. Midbody scale-rows 25-27; ventrals 211-230; subcaudals 
47-63; labials 7, the 3rd and 4th entering the orbit. 

Coloration. All four are of the rhombic type but the alleged model, 
Causus rhombeatus has not been recorded from any of these localities. 
It is interesting to note, however, that the young Egg-eater from 
Mwaya was thought to be the young of Vipcra superciliaris by the 
Banyakusa. 

Measurements. The largest, a female from Ukerewe Island, measures 
804 (720+84) mm., the smallest from Maji :Malulu, only 224 (190-f- 
34) mm. 

Diet. The first snake brought to me after our arrival on Ukerewe 
Island was an Egg-eater. I placed it in a vivarium and gave it a 
bronze Mannikin's (Spermestcs. c. scutatus) egg which it took; later 
two more eggs of the same species were removed from the stomach of 
a Boomslang {Dispholidus typiis) and though they were slightly 
cracked they were promptly taken. 

Enemies. A hedgehog {Atelerix a. hindei) was placed in the same 
cage for a couple of days and never molested the snake, on the third 
day, however, a larger and well-conditioned Egg-eater was temporarily 
placed in the same cage at noon but when I went to remove both 
snakes at sunset I found that the hedgehog had already nibbled a 
small hole in the belly of the still living snake and started to disem- 
bowel it. 

Habitat. The young Maji Malulu snake was found coiled beneath 
the skirting of the tent; when found it struck out open-jawed and kept 
its mouth wide-open all the time until picked up. I have never known 
an adult Egg-eater menace its would-be captor for thej' are among the 
most docile of snakes. 

BOIGINAE 

Tarbophis semiannulatus (Smith) 

Telescopus semiannulatus A. Smith, 1849, Illus. Zool. S. Africa, 3, pi. 72: No 
locality given. (South Africa.) 

cf (M. C. Z. 30209) Bagamoyo. 14. xi. 29. 

Distribution. Recorded by Sternfeld from Dar es Salaam, Ugogo, 
Lake Nyasa, Tukuyu and Lake Tanganyika. 

Variation. Midbody scale-rows 19; ventrals 212; anal divided; 
subcaudals 74; labials 8, the 3rd, 4th and 5th entering the orbit. 

Measurements. Total length of this cf 555 (450-|-105) mm. 



loveridge: African herpetology 247 

Diet. A Palm Gecko (Phclsuma dubium) was recovered from its 
stomach. 

Parasites. Larval acanthocephalans were encysted on the intestinal 
wall. 

Habitat. Taken at a height of eight feet from the ground in a young 
coconut palm. 

Crotaphopeltis HOTAiMBOEiA HOTAMBOEiA (Laurenti) 

Coronella hotamboeia Laurenti, 1768, Syn. Rept., p. 85: India oriental!, i.e. 
Africa. 

1 (M. C. Z. 30210) Unyanganyi, Turu. 4. xi. 29. 
22 (M. C. Z. 30211-32) Mwaya, Lake Nyasa. 1-8. iii. 30. 

4 (M. C. Z. 30233-6) Kitungulu, Urungu. 15. v. 30. 

5 (M. C. Z. 30237-41) Ujiji, Lake Tanganyika. 28. v. 30. 
1 (M. C. Z. 30242) Lkerewe Id., Lake Victoria. 12. vi. 30. 
1 (M. C. Z. 30243) Entebbe, Uganda. 28. vi. 30. 

Distribution. Recorded from Bukoba by Roux. 

Native names. Kiko (Kinyakusa); tukom'pe (Kirungu); ? sicela 
(Kijiji, but probably error as this is the Kinyamw^ezi name for Naja 
nigricollis which is also black). 

Corrigenda. Under the heading "Variation" in Barbour & Loveridge, 
1928, Mem. Mus. Comp. Zool. 50, p. 125, the word "not" was dropped 
out in the typing of the paper where it should read "preocular not in 
contact with the frontal." 

Variation. Midbody scale-rows 19; ventrals 156-172; anal single; 
subcaudals 29-46; labials 8, the 3rd, 4th and 5th entering the orbit 
except in Xo. 30229 where there are 9 with 4th, 5th and 6th entering, 
and Xo. 30220 where this condition, due to a division of the 2nd labial, 
occurs on one side of the head only; 1 preocular not in contact with 
the frontal except in Xo. 30233 and on right side of No. 30222; post- 
oculars 2, except X'o. 30214 which has 3; temporals 1+2 except in 
Xos. 30224 and 30233 where 1 + 1, due to the division of the anterior 
temporal, occurs on the left side of the head only; twelve snakes have 
3 pairs of chin-shields, three have a 3-4 arrangement, eighteen have 4 
pairs, and one (Xo. 30238) has 5. 

Coloration. In the Ukerewe Island snake the throat is deep black 
while in X^'o. 30219 from Mwaya the underside of the tail is black. 

Measurements. The largest male measures 573 (500+73) mm. and 
is from Ujiji, the biggest female is from Mwaya and measures 570 
(500+70) mm. 



248 bulletin: museum of comparative zoology 

Breeding. Eight of the Mwaya snakes are under 210 mm., the 
smallest measuring 150 (130+20) mm., the whole Kitungulu series 
are also young being under 242 mm. in total length. 

Diet. Stomach contents consisted of. — (1) A mouse (? Leggada 
hello) at Mwaya, (2) a toad {Bvfo r. regularis) at Ujiji, (3) Small frogs 
{Arihroleptis xenodactylus) in each of the four snakes from Kitungulu. 

Habitat. I took all the Kitungulu snakes in the course of a morning. 
They were beneath the bark, or in cavities, of fallen logs on the edge 
of dry forest. 

Crotaphopeltis hotamboeia tornieri (Werner) 

Leptodira tornieri Werner, 1876, Sitzber. Akad. Wien, 116, p. 1875: Amani, 

Usambara Mtns., Tanganyika Territory. 
Crotaphopeltis hotamboeia tornieri Barbour & Loveridge, 1928, Mem. Mus. 

Comp. Zool., 50, pp. 126-128: Usambara and Uluguru Mtns. — many 

localities. 

2 (M. C. Z. 30244-5) Dabaga, Uzungwe Mtns. 1. i. 30. 
1 (M. C. Z. 30246) Kigogo, Uzungwe Mtns. 13. i. 30. 
7 (M. C. Z. 30247-53) Madehani, Ukinga Mtns. 14-28. ii. 30. 
21 (M. C. Z. 30254-74) Nkuka Forest, Rungwe Mtn. iii. 30. 

Distribution. Hitherto only known from the Usambara and Uluguru 
Mountains. 

Native names. Nyamweru (Kihehe); nyoka naliombo (Kikinga). 

Corrigenda. On line 1 and line 15 of the 1928 citation given above 
for "3 preoculars" read 2 preoculars; it is correct but blurred in the 
typescript but was altered in its passage through the press. 

Variation. Midbody scale-rows 17; ventrals 162-175; anal single; 
subcaudals 35-48; labials usually 8, the 3rd, 4th and 5th entering the 
orbit, but No. 30257 has 7 with 3rd and 4th entering, three snakes 
have 8 with 4th and 5th, two snakes have 9 with 4th, 5th and 6th 
while three others ha^•e this arrangement on one side of the head only, 
the other side being normal; 2 preoculars in twenty-seven snakes, 1 
preocular in No. 30257 while three display an azygous arrangement 
with 2 on one side and 1 on the other, the preocular is not in contact 
with the frontal in eighteen snakes and is in contact in thirteen; 
postoculars 2 in twenty-one snakes, 3 in five and an azygous combina- 
tion in five; temporals l-f-2 in all but two snakes where they are l-f-l 
and No. 30264 which has an azygous combination. 

It will be seen that the differences between C. h. hotamboeia and 
€. h. tornieri are more sharply emphasized in this fresh material than 



loveridge: African herpetology 249 

they were in those from the Uluguru and Usambara Mountains and ad- 
jacent regions. Combining the information now available the two 
races may be distinguished as follows: — 

Scales in 19-21 rows (17 in one from the Cape 
fide Boulenger); ventrals 144-180; subcaudals 
29-54; preocular 1 (rarely 2); postoculars 2 
(rarely 1). Habitat outside rain forest, chiefly at 

low altitudes C.h. hotamboeia 

Scales in 17 rows (rarely 19, i.e. 10% of 76 snakes) ; 
ventrals 145-175; subcaudals 35-56; preoculars 
2 (rarely 1, i.e. in 21% of 76 snakes) ; postoculars 
2 (rarely 3). Habitat always in rain forest at 

high altitudes C.h. tornieri 

In life the forms are more readily separable than in the laboratory. 
Coloration. Yornig snakes are white beneath, while the adults are 
usually dusky, occasional specimens are as deep plumbeous below as 
above. The following notes were made in the field: Kigogo. Above, 
iridescent blue-black; a portion of the upper labials cream-colored. 
Below, throat white; ventrals cream, edged laterally with smoky- 
grey; subcaudals very pale smoky-grey. Eye red with blue-black 
vertical pupil. Madehani. Above, iridescent olive on head, plumbeus 
or blue-grey on body; upper labials, mental and four anterior lower 
labials, light yellow. Below, throat white; ventrals greenish-white. 
Eye orange-red with vertical black pupil. 

Measurements. The largest male (Rungwe) measures 581 (505+76+ 
tip) mm., and the biggest female (Dabaga) measures 545 (500+45+ 
tip) mm. It is unfortunate that both the largest snakes have lost the 
ends of their tails but no others are over 500 mm. in length from snout 
to anus. 

Breeding. The smallest snake measures 145 (125+20) mm. and 
was taken on March 31, 1930 at Rungwe. 

Diet. At Madehani each of three snakes had swallowed a frog 
{Arthroleptis reichei), the feet of a frog (Probreviceps m. rungwensis) 
were recovered from a snake taken in the Nkuka Forest. 

Enemies. Six snakes or nineteen per cent of the series have lost 
the end of their tails. 

Habitat. All of these snakes were taken in, or on the edge of, the 
rain forest, generally coiled beneath the bark of logs or beneath the 
trunks of fallen trees. Xone attempted to bite when caught. 



250 bulletin: museum of comparative zoology 



Amplorhinus nototaenia (Giinther) 

Coronella nototaenia Giinther, 1864, Proc. Zool. Soc. London, p. 309, pi. 26; 
fig. 1 : Rios de Sena, Zambesi. 

9 (M. C. Z. 30276) Nyamkolo, Lake Tanganyika. 9. v. 30. 
9 (M. C. Z. 30277) Kitungulu, Urungu. 14. v. 30. 

Variation. Midbody scale-rows 17; ventrals 167-161; anal divided; 
subcaudals 79-73 ; labials 8, the 4th and 5th entering the orbit though 
on the left side of the head of the Nyamkolo snake ; a wedge-shaped 
scale is interposed between the 2nd and 3rd labials and narrowly 
borders the lip but has been omitted from the count which is otherwise 
normal. 

It might be remarked here that Amplorhinus taeniahis Sternfeld 
from Lamu Island is a synonym of Hcmirhagerrhis kelleri Boettger; 
it is surprising that Boulenger did not detect this. 

Diet. The tail of a gecko (Lygodactylus ? stevensoni) was found in 
the stomach of the Kitungulu snake. 

Habitat. The Kitungulu snake was found resting on the horizontal 
branch of a fallen tree in drv miombo forest. 



Trimerorhinus tritaeniatus tritaeniatus (Giinther) 

Rhagerrhis tritaeniatus Giinther, 1863, Ann. Mag. Nat. Hist. (4), 1, p. 423, 
pi. xix, fig. 8: South East Africa. 

10 (M. C. Z. 30278-87) Dabaga, Uzungwe Mtns. 1-3. i. 30. 

1 (M. C. Z. 30288) Ihanganya, Uzungwe Mtns. 6. i. 30. 

1 (M. C. Z. 30289) Ipemi, Uzungwe Mtns. 7. i. 30. 

7 (M. C. Z. 30290-6) Kigogo, Uzungwe Mtns. 13-30 i. 30. 

1 (M. C. Z. 30297) Lukungu, Ubena Mtns. 8. ii. 30. 

1 (M. C. Z. 30298) Ihenye, Ukinga Mtns. 8. ii. 30. 

3 (M. C. Z. 30299-301) Mangoto, Ukinga Mtns. 10. ii. 30. 

1 (M. C. Z. 30302) Ilolo, Rungwe district. 22. iii. 30. 
10 (M. C. Z. 30303-12) Igale, Poroto Mtns. 24-30. iv. 30. 

Distribution. iVlso, according to the Wakinga, abundant in the 
grasslands at Madehani during the heaviest rains. Sternfeld has re- 
corded T. rhombeatiis as occurring in these Ukinga Mountains but it 
probably came from the southern end of the range while Madehani 
is in the extreme north. 

Native navies. Nyaluhereka (Kikinga) ; isakani (Kinyakusa, but not 
even generic as applied to two other striped snakes). 



loveridge: African herpetology 251 

Variation. Midbody scale-rows 17; ventrals 145-162; suboaudals 
51-60; labials 8, the 4th and 5th entering the eye; rostral broader 
than deep as in the s^^lonym variabilis from the Shire highlands. In 
four snakes only is it as broad as deep. 

Coloration. Very young snakes differ from the adults in appearance 
as the upper labials are white and though their longitudinal stripes are 
not more numerous they appear to be so as a result of their being in 
closer juxtaposition in the small reptiles. 

Measurements. The largest snake, a female from Igale, measures 
993 (827+146) mm. but has the tip of the tail missing; the smallest 
snake is from Dabaga and measures 186 (148+38) mm. 

Sex. Females appear to predominate; in the Igale series for example 
only two of the ten snakes are males. 

Diet. Stomach contents consisted of: — (1) Large shrew (Crocidura 
nyanzae kivu) at Kigogo, (2) rat remains at Igale, (3) skinks (Mabuya 
V. varia) in each of five snakes at Dabaga, Kigogo and Igale, (4) hind 
limbs of frog (Rana sp.) at Dabaga, (5) frog {Rana f. angolcnsis) at 
Kigogo. 

Parasites. Nematodes {Kalicephalus sp.) were recovered from the 
stomach of Dabaga and Ihenye snakes. 

Defence. These snakes emit a cloacal discharge similar to that of 
the European Grass-Snake {Natrix n. natrix), the smell being in- 
distinguishable to me, though the quantity of the offensively odorifer- 
ous fluid was less. 

Temperament. Instead of seeking safety in swift flight like its 
relatives of the genus Psammophis, the Striped Schaapsteker makes 
for the nearest grass-enveloped shrub and conceals itself at the base, 
defying all efforts to drive it out; often relying so much on remaining 
quiet that it may be picked up with ease. Once seized, however, it 
will thrash about, flatten its body to a surprising extent and some- 
times even bite, though this is unusual. The bite, while drawing blood, 
is not followed by symptoms of poisoning. 

Enemies. Fifty per cent of the Igale series have lost the extreme end 
of their tails while those of the Uzungwe Mountains series are intact. 

Habitat. Young snakes were seen in fresh green grass bordering a 
marsh in a valley at Dabaga. Both the Ihanganya and Ipemi snakes 
shot across the path between porters who were on the march, the 
porters killed one and I caught the other. When clearing the site for 
my tent at Igale one of these snakes was disturbed from the base of 
a grass-smothered shrub; I caught a second a few minutes later within 
ten feet of the tent; I captured two more within a couple of hundred 



252 bulletin: museum of comparative zoology 

yards and all were taken in less than an hour. Five more were brought 
in by the natives engaged in building a grass hut for the skinners. 

Rhamphiophis acutus (Giinther) 

Psavmrophis acutus Giinther, 1888, Ann. Mag. Nat. Hist. (6), 1, p. 327, pi. 
xix, fig. D: Pungo Andongo, Angola. 

? (M. C. Z. 30313) Mwaya, Lake Nyasa. 1-8. iii. 30. 

Distribution. This record adds a species to the snake fauna of 
Tanganyika Territory, for hitherto 7?. acutus has been known only from 
Angola. 

Affinities. Perhaps the most important generic character separating 
Trimerorhinus from RhaviphiojjJiis has been the possession of 10 to 12 
maxillary teeth plus two fangs by the former and 6 to 9 plus 2 fangs 
by the latter; the specimen before me, however, has 10 plus 2 fangs 
and on each side anteriorly are two azygous gaps which seem to indi- 
cate that the full complement of teeth is 12 plus 2 fangs. The peculiar 
rostral, hollowed out beneath, remains to separate the genera. 

I might add that the coloring of acutus is line for line and in every 
detail identical with that of typical T. triiaeniatus except that the 
lower surface of acutus is white and that of tritaeniatus is dark except 
for occasional pale specimens. This is a further indication of the very 
close relationship between the two reptiles, acutus appearing to form 
a connecting link between the genera. 

Variation. It differs from the type in that each loreal has a portion 
split off the lower part and the preocular is narrowly separated from 
the frontal; in number of ^'entrals and subcaudals and other scale 
formulae it is identical with the type though the former was a male. 

Measurements. Total length 825 (670+155) mm. 

Habitat. Taken in hot low -lying country at an altitude of 1,700 feet 
above sea level ; this is in sharp contrast to the habitat of T. tritaeniatus 
which favours the grasslands of the high plateaux at 5,000 feet and 
over. 

Rhamphiophis oxyrhynchus (Reinhardt) 

Psammophis oxyrhynchus Reinhardt, 1843, Dansk. Vidensk. Selsk. Skrift., 10, 

p. 244, pi. i, figs. 10 and 12: Guinea, West Africa. 
Rhamphiophis connali Parker, 1929, Ann. Mag. Nat. Hist, (10), 4, p. 449: 

Accra, Gold Coast. 

Parker has recently proved conclusively that Boulenger was in 
error in considering that West and East African snakes were referable 



LOVERIDGE: AFRICAN HERPETOLOGY 253 

to a single species for which Boulenger employed the name Rhmn- 
phiophis o.vyrln/nchus in the Catalogue of Snakes, 3, p. 146. Having 
done this Parker proposed the name R. connali for West African 
snakes from the Gold Coast, Dahomey and Nigeria. Unfortunately 
Reinhardt's type came from Guinea, West Africa and- whether we 
regard that locality in its present restricted sense or in the vague 
Gambia to Gaboon fashion prevalent a century ago, makes little 
difference, for the fact that Reinhardt's snake was of West African 
origin is proved by his statement that it had a single preocular. 

Another name must be sought for East African snakes and the 
earliest available appears to be RhainphiopMs rostratus described by 
Peters in 1854 and of whose identity there can be no doubt, for a fine 
colored plate was published shortly after the original description. 

Rhamphiophis rostratus Peters 

Rhamphiophis rostratus Peters, 1854, Monatsb. Akad. Wiss. Berlin, p. 624; 
1882, Reise nach Mossamb., 3, p. 124, pi. xix, fig. 1: Tete; Mesuril; 
Quitangonha, Mozambique. 

1 (M. C. Z. 30314) Dar es Salaam. 8. xi. 29. 

2 (M. C. Z. 30315-6) Bagamoyo. 16. xi. 29. 

6 (M. C. Z. 30317-22) Mangasini, Usandawi. 16. xii. 29. 

Distribution. Recorded by Sternfeld from Dar es Salaam, Mpwapwa, 
Bukoba etc. 

Affinities. The name rostratus has had to be revived for all East 
African snakes hitherto known as oxyrhynchus since Parker has shown 
that West Coast snakes differ in possessing a single preocular and 
a preanal scale count that never falls below 15 rows. 

Variation. Midbody scale-rows 17; preanal scale-rows 13; ventrals 
165-192; anal divided ;subcaudals 104-113; labials 8, the 4th and 5th 
entering the eye except in three snakes where only the 5th enters; in 
these snakes, and one other, there are 3 preoculars, the remaining five 
have 2 preoculars, in no specimen are these in contact with the frontal ; 
anterior chin -shields are as long as the posterior except in Nos. 30314-5 
where they are shorter. 

Measurements. None is of exceptionally large size, the biggest 
(No. 30317) being only 1,335 (930+405) mm. 

Diet. Stomach contents consisted of: — (1) Lizard {Eremias s. 
spekii) at Mangasini, (2) skink {Riopa s. sundevallii) and three frogs 
(Arthroleptis s. stenodactylus) in the Dar es Salaam snake, (3) an elater 
beetle in the fully adult Bagamoyo reptile. 



254 bulletin: museum of comparative zoology 

Dromophis lineatus (Dumeril & Bibron) 

Dryophylax lineatus Dumeril & Bibron, 1854, Erpet. Gen., 7, p. 1124: White 
Nile, Africa. 

3 (M..C. Z. 30323-5) Mwaya, Lake Nyasa. 1-8. iii. 30. 
1 (M. C. Z. 30326) Nyamkolo, Lake Tanganyika. 9. v. 30. 
1 (M. C. Z. 30327) Ujiji, Lake Tanganyika. 29. v. 30. 

Distribution. Previously recorded from Tukuyu and Ipiani, near 
Mwaya, by Sternfeld. 

Native name. Isakani (Kinyakusa, but not even generic). 

Variation. Midbody scale-rows 17; ventrals 146-154; anal divided; 
subcaudals 83-85; labials 8, the 4tli and 5th entering the eye; preocular 
1, not in contact with the frontal; postoculars 2 except in the Ujiji 
snake which has a single postocular on the right side; temporals 
normal in two snakes only, i.e. 1+2, azygous in the other three as 
1+2, 1+3, 2+2 and 2+3. Boulenger in his 1915 key differentiates 
the genus Dromophis from Psammioplns on the grounds that the 
former has only a single temporal and the latter two but the Nyamkolo 
snake has 2 on both sides of its head and a Mwaya snake on one side 
only. 

Coloration. Though easy to recognise in life, when preserved this 
species is strikingly like Psammophis suhtaeniatus , all five specimens, 
however, have the transverse lateral streaks on the ventrals while 
P. suhtaeniatus usually has longitudinal lines but never transverse. 

Measurements. All are females, the largest being 861 (600+261) 
mm., and is from Mwaya. 

Breeding. The Ujiji snake held six eggs each measuring 15 x 6 mm. 

Diet. A frog {Rana m. mascareniensis) was found in the Ujiji snake. 

Psammophis subtaeniatus Peters 

Psammophis sibilans var. subtaeniata Peters, 1882, Reise nach Mossamb., 3, 
p. 121: Boror and inland from Tete, Mozambique. 

5 (M. C. Z. 30328-31, 30339) Unyanganyi, Turu. 4. xii. 29. 
2 (M. C. Z. 30337-8) Saranda, Ugogo. 29. xi. & 18. xii. 29. 
21 (M. C. Z. 30340-60) Mangasini, Usandawi. 12-16. xii. 29. 
2 (M. C. Z. 30332-3) Mwanza, Lake Victoria. 6. vi. 30. 
1 (M. C. Z. 30370) Ukerewe Island, Lake Victoria. 12. vi. 30. 

Distribution. Another was seen crossing the road ten miles south of 
Bagamoyo. The species has been recorded by Sternfeld for Kitopeni; 
Lake Nyasa; Tukuju and Ukerewe Island. 



loveridge: African herpetology 255 

Native name. Nne, followed by a click (Kisandawi). 

Variation. jNIidbody scale-rows 17; ventrals 153-176; anal divided; 
subcaudals 86-115; preocular 1, rarely 2; temporals very variable 1 
over 2+2 or 3, 2+2, 2+3; rostral broader than deep in twenty-five 
snakes, as broad as deep in six demonstrating that this character, 
utilized by Boiilenger in his 1915 key to distinguish sithtacniatus from 
sibilans is useless. Though they are good and full species, I am at a 
loss to differentiate these two snakes by anything but color. Stern- 
feld reached the same conclusion in 1908. 

Coloration. The pair of longitudinal lines along the whole under- 
surface serve to distinguish this species from sibilans, the Saranda, 
Mangasini and Ukerewe Island snakes are by no means typical for 
the lines, instead of being sharply defined are dusky, sometimes very 
faint and in others broken up into a series of dashes, the line usually 
separates the cream-colored belly from the ventro-lateral band of 
white. The bigger snakes from these localities were so similar to 
sibilans in their dorsal coloration that I considered them to be referable 
to that species in the field. 

Measurements. In size also these large Mangasini snakes surpassed 
the largest subtaeniatus with which I was acquainted and forty-six 
were secured on a former tour. The biggest snake (No. 30340) in the 
present series measured 1,370 (1,030+340) mm., the smallest (No. 
30338) was 332 (225+107) mm. 

Diet. One Mangasini snake held a rat {Rattus r. kijabius), another 
a lizard {Eremias s. spekii) while each of two Unyanganyi reptiles had 
swallowed a lizard {Nucras b. boulengeri). 

Parasites. Nematodes ( 9 , and immature Spiuroidea) were recovered 
from the Mangasini and Ukerewe Island snakes. 

PsAMMOPHis SIBILANS (Linnaeus) 
Coluber sibilans Linnaeus (part), 1766, Syst. Nat., 12th ed., 1, p. 383: "Asia.' 

1 (M. C. Z. 30334) Dar es Salaam. 8. xi. 29. 

2 (M. C. Z. 30335-^) Bagamoyo. 14. xi. 29. 

4 (M. C. Z. 30361-4) Mwaya, Lake Nyasa. 1-8. iii. 30. 

1 (M. C. Z. 30365) Igale, Poroto Mtns. 30. iv. 30. 

2 (M. C. Z. 30366-7) Near Ikombo, N. Rhodesia. 6. v. 30. 
1 (M. C. Z. 30368) Nyamkolo, Lake Tanganyika. 9. v. 30. 

1 (M. C. Z. 30369) Kalambo River, Lake Tanganyika. 12. v. 30. 

Distribution. Also recorded by Sternfeld from Bagamoyo and 
Ukerewe Island, and by Roux from Bukoba. 



256 bulletin: museum of comparative zoology 

Native names. Ngaruka (Kinyakusa); mlalu (Kiningu). 

Variation. Midbody scale-rows 17; ventrals 159-179; anal divided; 
subcaudals 86-100; rostral broader than deep in nine snakes, as broad 
as deep in three; see remarks regarding the specific value of this 
character under P. subtaeniatus . 

Coloration. Both Bagamoyo snakes are very young, one has fine 
speckling on the ventral scutes while the other has these scales un- 
spotted. 

Measurements. The largest snake (No. 30362) is 1,245 (1,000+245) 
mm. in length but lacks the end of its tail; the smallest snake (No. 
30336) is 310 (220+90) mm. 

Diet. Stomachcontents consisted of: — (1) Skink (Mabuya viaculila- 
bris) at ]Mwa\-a, (2) frog {Arthroleptis minutus) at Mwaya, (3) frog 
{Rana m. mascarcnicnsis) at Nyamkolo. 

Parasites. Nematode worms {Physalopteraparadoxa and Kalicephalvs 
sp.) were found in the stomach of a Mwaya snake; indeterminate 
nematodes in a Nyamkolo specimen. 

Habitat. One of the juvenile snakes from Bagamoyo was found in 
the market. Our lorry ran over the large Hissing Sand Snake which 
was crossing the road near Ikombo; later the same morning I captured 
the smaller one which was ensconced on the dashboard, presumably it 
had been swept on to the car from the dense grass and brush through 
which we had been driving. The reptile listed from Kalambo River 
was in a hollow gourd lying on a sandy waste near the mouth of the 
Kalambo River but on the southern, i.e. North Rhodesian, bank. 

Psammophis biseriatus Peters 

Psammophis biseriatus Peters, 1881, Sitzb. Ges. Naturf. Freunde Berlin, p. 88: 
Taita, Kenya Colony. 

1 (M. C. Z. 30371) Kilimatinde, Ugogo. 26. xi. 29. 

2 (M. C. Z. 30372-3) Unyanganyi, Turu. 4. xii. 29. 

9 (M. C. Z. 30374-82) Mangasini, Usandawi. 12-16. xii. 29. 
4 (M. C. Z. 30383-6) Kikuyu, Ugogo. 21. xii. 29. 

Xatire name. Kutlaku, with a click (Kisandawi). 

Variation.. Midbody scale-rows 15; ventrals 150-155; anal divided; 
subcaudals 97-114; labials 9, the 4th, 5th and 6th entering the eye; 
temporals azygous in four snakes so that twenty sides have 2+3, 
eleven have 2+2 and one is 1+2 in defiance of the key in the Cata- 
logue of Snakes, in two specimens the lower of the anterior pair of 
temporals has been subdi\ided vertically. 



loveridge: African herpetology 257 

Measurements. The largest snake, a female from Mangasini, 
measures 865 (565+300) mm.; the smallest is from Unyanganyi and 
measures 300 (200+100) mm. but six snakes taken early in December 
are 325 mm. 

Diet. A Two-striped Sand Snake was disturbed on the path near 
Maji Malulu on 10. xii. 29 as it was about to swallow a lizard (Nucras 
b. boulcngeri) whose head was already in its mouth. As I jumped from 
my cycle and attempted to grab the snake the latter made off at great 
speed carrying the lizard, I eventually lost trace of it in a tangle of 
fallen thorn bush. 

Skinks {Riopa simdevallii modcstum) were recovered from the 
stomachs of two of the Mangasini reptiles. 

Parasites. An indeterminate female ascarid was taken from the anus 
of one Mangasini snake. 

Habitat. At Dodoma a young one was seen near a manyara hedge at 
sunset. At Kilimatinde a young one was taken at 4 p.m. among rubble 
in the old fort ; just after sunset the same evening an adult darted into 
a pile of river debris in the river bed. At Mangasini I was returning 
to camp in a rainstorm when one of these snakes crossed the path 
and ascended a thorn tree in which I caught it; this occurred at 6 p.m. 
or about half-an-hour before dark. Driving from Dodoma to Iringa 
I saw two on the road but both within ten miles of Dodoma. 

PSAMMOPHIS ANGOLENSIS (Bocagc) 

AmpMophis angolensis Bocage, 1872, Jorn. Sci. Lisboa, 4, p. 82: Donda, 
interior of Angola. 

1 (M. C. Z. 30387) Unyanganyi, Turu. 3. xii. 29. 

Distrilndion. Recorded by Boulenger from Lakes Nyasa, Tangan- 
yika and Victoria. 

Variation.. The example of this rare little snake is wholly normal 
with midbody scale-rows 1 1 ; ventrals 149 ; subcaudals 64 and answer- 
ing in all respects to the revised description in the Catalogue of Snakes, 
3, p. 170. 

Measurements . Total length 328 (245+83) mm. 

Habitat. Taken in an mbugwe or dried-up flat. 

ThELOTORNIS KIRTLANDII (Hallowcll) 

Leptophis kirtlandii Hallowell, 1844, Proc. Acad. Nat. Sci. Phila., p. 62: Liberia. 

c? (M. C. Z. 30388) Kitungulu, Urungu. 15. v. 30. 



258 bulletin: museum of comparative zoology 

Distribution. Sternfeld lists the Bird Snake from Mpwapwa, Lake 
Xyasa and Lake Mctoria. I was showTi a large example at Ilolo. 

Native names. Lukukuru (Kikami); lukungu (Kinyika); ncmdo 
(Kirungu); nalakutu (Kij^ao). 

Variation. Midbody scale-rows 19; ventrals 159; subcaudals 145. 

Measurements. Total length 1,140 (690+450) mm. 

Habitat. This snake was obtained under rather unusual circum- 
stances. There was a huge heap of dry grass and rubbish surrounding 
the base of two great trees in a native clearing. I had the heap ignited 
as it seemed probable that there would be cobras in such an ideal spot. 
The heat from the blaze rose into the trees though the flames fell far 
short; towards the end of the conflagration this Bird Snake dropped 
from a height of at least twenty feet. Though I saw it fall I mistook 
it for a branch till a shout from one of the "boys" drew my attention 
to the departing snake which I pursued and captured among the 
standing maize. 

DisPHOLiDUS TYPUS (Smith) 

Bucephalus typus A. Smith, 1829, Zool. Journ., 4, p. 441 : Old Latakoo, South 
Africa. 

1 (M. C. Z. 30389) Mpwapwa, Ugogo. 22. xi. 29. 

2 (M. C. Z. 30390-1) Unyanganyi, Turu. 6. xii. 29. 
1 (M. C. Z. 30392) Masiliwa, Turu. 10. xii. 29. 

3 (M. C. Z. 30393-5) Mangasini, Usandawi. 14. xu. 29. 
1 (M. C. Z. 30396) Mwaya, Lake Nyasa. 1-8. iii. 30. 

1 (M. C. Z. 30397) Ukerewe Id., Lake Victoria. 12. vi. 30. 
1 (M. C. Z. 30398) Kampala, Uganda, vi. 30. 

Distribution. Sternfeld has aheady recorded the Boomslang as 
occurring at Mpwapwa and L^kerewe Island, also at Dar es Salaam 
and Bukoba. 

Native name. Imbindipindi (Kinyakusa, but applied to all green 
tree snakes) . 

Variation. Midbody scale-rows 19; ventrals 171-188; anal divided; 
subcaudals 95-111 ; labials 7, the 3rd and 4th entering the orbit except 
on the left side of No. 30390 where there are 8 labials with 4th and 5th 
entering the orbit; preocular 1; postoculars 3; temporals 1+2 except 
in No. 30396 where they are 2+2 and No. 30390 where they are 2+3 
on the right side and 1+3 on the left side of the head. 

Coloration. The jNIpwapwa and Ukerewe Island snakes are al- 
together green; the JNIwaya, Kampala and one Mangasini snakes are 



loveridge: africax herpetology 259 

green with black markings; the largest of the three Mangasini snakes 
was olive-color and extraordinarily like a Mamba {Dendraspis angus- 
ticeps); the medium-sized one was vinous with white labials and suffi- 
ciently like a Bird Snake {Thelotornis kirtlandii) for me to mistake it 
for that species in the field; the two young Unyanganyi snakes were 
noted in life as "Above, dark brown with pale blue speckling particu- 
larly conspicuous on the neck which is otherwise black, the skin in this 
region which is shown when the neck is inflated, is of the same shade 
of pale blue; upper labials and throat white with a patch of pale yellow 
at the base of the jaws; the rest of the lower surface brownish grey 
with dark broAMi mottlings." 

Measurements. The largest specimen, though at least a hundred 
millimetres of its tail are lacking, is a male from Mangasini which 
measures 1,495 (1,210+285+tip) mm. 

Diet. Mr. Evans of the Veterinary Station, Mpwapwa described a 
large green snake to me which was almost certainly a Boomslang. It 
lived in the roof of his house and from time to time raided the swallows' 
nests from which he had seen it carry off the young. 

I was cycling along a path near Masiliwa when I observed an olive- 
green Boomslang on the ground swallowing a frog {Leptopelis hocagii). 
As I was about to seize the snake it disgorged its prey; later I found a 
second specimen and a partly -digested Breviccps mossambicus in its 
stomach. 

A Mangasini snake had eaten a chameleon {Chamaeleon d. dilepis). 

While I was sitting writing in my tent I heard a slight commotion 
among the dry leaves beneath a mango tree a hundred feet away and 
near the mission station on Ukerewe Island. A native was jumping 
about very actively and slashing the ground with his stick, now here, 
now there. Running to the spot I was in time to capture a vivid green 
Boomslang before he annihilated it. The native said that he was walk- 
ing beneath the tree when the snake fell down close behind him. In 
its stomach were a clutch of eggs of the Ethiopean Bronze Mannakin 
(Spermestes c. scutatus) a species which had several nests in the foliage 
of the mango where I had observed the birds, some of the eggs were 
unbroken, the others may have been broken by blows from the native's 
stick. 

Parasites. A nematode, which was not preserved, was found in the 
stomach of the last-mentioned snake. 

Defence. The Mpwapwa snake, quite a large one, was disturbed by 
Salimu and came straight down the ravine towards me. I pinned it 
as it would have passed, but only by the tail, whereupon it raised itself 



260 bulletin: museum of comparative zoology 

high in the air and menaced me with open jaws and inflated throat, its 
head was level with my face for it was higher up the bank than I was. 
Having nowhere else to keep it I put it in an empty kerosene tin for 
twenty-four hours, on removing it I found the creature was quite 
stupified, presumably by the smell for the tin had been placed in a 
cool stone room. 

Calamelaps unicolor (Reinhardt) 

Calamaria unicolor Reinhardt, 1843, Dansk Vidensk. Selsk. Skrift., 10, p. 236, 

pi. i, figs. 1-3: Guinea, West Africa. 
Calamelaps polylepis Bocage, 1873, Jorn. Sci. Lisboa, 4, p. 216: Dondo, interior 

of Angola. 
Calamelaps miolepis Giinther, 1888, Ann. Mag. Nat. Hist, (6), 1, p. 323: Cape 

McCIear, Lake Nyasa. 
Calamelaps unicolor Barbour & Loveridge, 1928, Mem. Mus. Comp. Zool., 50, 

p. 130: Uluguru and Usambara Mtns., Tanganyika Territory. 

1 (M. C. Z. 30399) Mwaya, Lake Nyasa. 1-8. iii. 30. 

Distribution. Recorded by Sternfeld from Bagamoyo but he has 
overlooked Tornier's 1901 paper on specimens from the Tanga- 
Usambara region. 

Native name. Nyeresi (Kinyakusa). 

Affinities. The Mwaya snake is almost a topotype of miolepis which 
Boulenger relegated to the synonymy of poh/lepis in 1896. The 
reasons for considering polylepis itself a synonym of unicolor will be 
found in the 1928 citation given above. It has since occurred to me 
that polylepis may stand in the same relation to unicolor as does 
Crofaphopeltis h. hotamhocia to C. h. tornieri, i.e. that snakes with 17 
raidbody scale-rows may be restricted to the tropical or mountain 
rain-forest areas while those with 19 or 21 scale-rows occur only in 
hot country at lower altitudes. In the main this is supported by the 
records but is negatived by Sternfeld's Bagamoyo specimen and some 
others. More material of this rare species is required before a mature 
decision can be reached. 

Ckdariuiaps mellandi, described by Boulenger in 1915 from a male 
from Chirini Island, Lake Bangweulu, was differentiated from polylepis 
by the absence of a postocular; 2nd upper labial in contact with the 
prefrontal; 4th upper labial forming a suture with the parietal; and 
181 ventrals. In passing I might say that of four Amani snakes one 
(No. 23359) has the 2nd upper labial in contact with the prefrontal 
and two have the 4th upper labial forming a suture with the parietal 



loveridge: African herpetology 261 

as is the case with the IMwaya snake and on one side of the head only 
in a Lumbo snake. C. mcUandi remains distinguished by the absence 
of a postocular, a character of doubtful specific value in a species 
where the postocular is already reduced to a mere vestige. 

Variation. Midbody scale-rows 19; ventrals 177; anal divided; 
subcaudals 28; labials 6, the 3rd and 4th entering the orbit, postocular 
1; temporal 1. 

Measurements. Total length 475 (430+45) mm. 



Rhinocalamus dimidiatus Giinther 

Rhinocalamus dimidiatus Giinther, 1888, Ann. Mag. Nat. Hist. (6), 1, p. 322, 
pi. xix, fig. C : Mpwapwa, Ugogo, Tanganyika Territory. 

1 (M. C. Z. 30400) Mpwapwa, Ugogo. 23. xi. 29. 

Distribution. Only known from the type locality. 

Variation. The scalation is normal. Midbody scale-rows 17; ven- 
trals 199; anal divided; subcaudals 26. 

Coloration in life. Above, glossy black, except the labials and a 
lateral band of chrome-yellow; below, white, except for the edges of 
the ventrals which are, like the tail, chrome-yellow. 

Measurements. Total length 400 (365+35) mm. 

Habitat. Taken six inches below the surface in a cavity of the 
rotting roots of a tree-stump which was situated in sand on the banks 
of the bed of a stream long since dried-up. 



MiODON GABONENSis (Dumeril) 

Elapomorphus gabonensis A. Dumeril, 1856, Rev. Mag. Zool. (2), 7, p. 468: 
Gaboon, West Africa. 

9 (M. C. Z. 30401) Ilolo, Rungwe. 15. iv. 30. 

Distribution. The only record in Africa east of Uganda of which I 
am cognisant is that of Sternfeld for Dar es Salaam which is also the 
only record for the occurrence of the genus in this region. 

Variation. Midbody scale-rows 15; ventrals 215; anal divided; sub- 
caudals 21 ; the frontal is once and a half times as broad as the supra- 
ocular; the internasals are as long as the prefrontals; the nasal is com- 
pletely divided though Boulenger states that it is entire or incom- 
pletely divided. The genus is badly in need of revision. 

Coloration. Uniformlv iridescent blue-black above and below. 



262 bulletin: museum of comparative zoology 

Measurements. Total length 479 (448+31) mm. 
Diet. What is unmistakably the tip of the tail of a blind snake 
(Typhlops or Leptotyphlops) was present in the stomach. 



Chilorhinophis gerardi (Boulenger) 

Aposiolepis gerardi Boulenger, 1913, Revue Zool. Afr., 3, p. 103, fig.: Kikondja, 

Katanga, Belgian Congo. 
Parkerophis gerardi Barbour & Amaral, 1927, Bull, Antivenin Inst. Amer., 1, 

p. 25; Parker, 1927, Ann. Mag. Nat. Hist. (9), 20, p. 81: Sinoia, Loma- 

gundi district, S. Rhodesia. 

c? (M. C. Z. 30402) Nyamkolo, Lake Tanganyika. 9. v. 30. 

Distribution. This record constitutes the first for Northern Rhodesia 
and is the third known specimen, since Boulenger's record from 
Anquabe, Portuguese East Africa has been shown by Parker to repre- 
sent a distinct species which he called Parkerophis carpenteri. I have 
little doubt that C. gerardi v:i\\ be found near Kasanga in Tanganyika 
Territory for the natives at Kasanga profess to know it under the first 
of the two names given below. 

Native names. Kasimwanamatengi, kalamhanzila (Kirungu). 

Affinities. The name Parkerophis proposed by Barbour and Amaral 
for gerardi must be considered a synonym of Chilorinophis, a genus 
erected by Werner for the reception of hutleri which he described from 
the Sudan in 1908 (1907) . Chilorhinophis hutleri has been recorded from 
Amani, Usambara Mountains, Tanganyika Territory by Sternfeld 
(1910). Assuming that the identification is correct it forms a link be- 
tween the Sudan record of hutleri and the ^Mozambique one of car- 
penteri. The ventral and subcaudal scale-counts of the two types are 
widely separated but supposing that they are of difl^erent sexes there 
is a remote possibility of their representing one species. 

Variation. Midbody scale-rows 15; ventrals 308; anal divided; 
subcaudals 26. The type had 276 ventrals and 28 subcaudals, in other 
respects they agree, I was also afforded the opportunity of direct 
comparison with the Sinoia snake in the British Museum. 

Coloration in life. The original description based on a preserved 
specimen gives one little idea of the beautiful appearance of the living 
reptile. Above, crown of head black flecked with yellow on the pre- 
frontals, supraoculars and parietals; it is also black for six scale-rows 
behind the head except for a pair of yellow flecks just posterior to the 
parietals; three black bands (a vertebral flanked by dorso-laterals) 



loveridge: African herpetology 263 

proceed from the black patch on the neck and are continued along the 
hody and tail until they merge into the black tip of the tail, this black 
tip is flecked with yellow like the head ; between the bands, and on the 
flanks, the body color is chrome yellow. Below, the throat is china- 
white extending upwards to some of the upper labials; the ventrals 
and a half scale-row on either side are orange as are also the anterior 
subcaudals followed by two pairs of black subcaudals, then ten pairs 
of white subcaudals with grey blue centres some of which are flecked 
with black, the terminal scute of the tail is black. 

Measurements. Total length 445 (420-f25) mm. 

Defence. I saw the reptile wriggling along with its tail held high 
simulating a head as described and figured by Carpenter for the allied 
form. 

Habitat. This snake was taken on the road leading up to the London 
^Missionary Society's station on the bluff overlooking the bay at 
Xj'amkolo, by men engaged in cleaning weeds off the road. The rains 
had ceased a month before and the country was already very dry. 
The villagers stated that the species was not rare at certain seasons 
but all my efforts to obtain others during the three days we camped at 
Xyamkolo failed. Curiously enough the veteran missionary Mr. White 
described this snake to me within an hour or so of my arrival saying 
that he had seen several twentv-five vears ago but none in recent 
times. 

ELAPIX.\E 

BOULENGERIXA ANNULATA STORMSI Dollo 

Boulengerina stormsi Dollo, 1886, Bull. Mus. Beige, 4, p. 160, fig: Lake Tan- 
ganyika; Boulenger, 1904, Ann. Mag. Nat. Hist. (7), 14, p. 15; Boulenger, 
1919, Revue Zool. Afr., 7, p. 27: Bosabangi, Ituri, Belgian Congo; Boulen- 
ger, 1919, Proc. Zool. Soc. London, p. 293: Key to species. 

9 (M. C. Z. 30403) Kasanga, Lake Tanganyika. 17. v. 30. 
o" (M. C. Z. 30404) Kipili, Lake Tanganyika. 19. v. 30. 

Native names. Miu (Kirungu); Mwiu (Kifipa); Mlolo (Kijiji). 

Distribution. Until 1919, when Boulenger referred a specimen from 
the Ituri to stormsi, it was believed that that species was confined to 
Lake Tanganyika. 

Affinities. Schmidt (1923, Bull. Am. Mus. Nat. Hist., 49, p. 123) 
has conclu3i^■ely shown that B. dyhoicshji Mocquard is synonymous 
with B. anmdata (Bucholtz & Peters) and concludes that stormsi may 
be distinguished by its 21 scale-rows (instead of 21-23 found in his 



264 bulletin: museum of comparative zoology 

series of anmdaia), a longer tail and different coloration. Boulenger 
has hinted that the two might not be specifically distinct. 

The two specimens listed above effectually dispose of the supposed 
differences in scale-rows and tail length so that we are left with only 
a color difference, well-marked in the extreme West and East of the 
range but intergrading in the Belgian Congo. This difference may be 
expressed as follows: — 

Western form with 20, or more, complete annuli on body a. annulata 
Eastern form with 2 (rarely 3 or none) annuli on body a. stormsi 

I am indebted to Mr. H. W. Parker for supplying me with the 
bulk of the data on which these conclusions are based ; it is listed below, 
together with the other data available to me. Localities are arranged 
from West to East and North to South. 

B. annulata annulata 

B. M. #6. Bitye, S. Cameroon. Adult skin. 

Black transverse occipital bar; 20 saddle-shaped light 
centered, black annuli on body; some indistinct annuli 
on tail. 
M. C. Z. 29358. Lukungg River, Bipindi, S. Cameroon. Adult. 

Black occipital bar; 5 solid annuli anteriorly on body fol- 
lowed by 15 light centered ones; indistinct annuli on tail. 
B. M. #8 & 9. Benito River, French Congo. Two juveniles. 

Black occipital bar; 3 solid annuli anteriorly on body fol- 
lowed by 19 light centered ones; indistinct annuli on tail. 
B. M. #7. Ogowe, French Congo. Ad. 

Black occipital bar; 20 light centered, black annuli on 
body; indistinct light centered ones on the tail. 
B. M. #11. Gaboon, French Congo. Juv. 

Black occipital bar; a black bar followed by 3 solid annuli 
and 18 light centered, black annuli on the body. 
B. M. #10. Ubangi, French Congo. Juv. 

No occipital bar; a black bar followed by 2 solid annuli 
and 21 light centered, black annuli on the body. 
Intermediate between the two races. (Referred to annulata by Schmidt.) 
M. C. Z. 13608. Ngayu, Belgian Congo. Ad. 

Black occipital bar; 2 solid black annuli followed by about 
(posteriorly they are very ill-defined) 18 light centered, 
white spotted, black bars on the body. 
B. annulata stormsi 
U. S. N. M. 63378. Ujiji, Lake Tanganyika. Juv. 

Black occipital bar; 2 solid black annuli followed by ten 
cross bars some of which show a light line through them, 
one or two are rather ill-defined. 



loveridge: African herpetology 265 

B. M. #2. Lake Tanganyika. Juv. 

Black occipital bar; 2 black annuli followed by 5 cross 
bars on the body. 
B. M. #1. Lake Tanganyika. Juv. 

Black occipital bar; 2 ^black annuli followed by 4 cross 
bars, becoming shorter posteriorly, on the body. 
M. C. Z. 30404. Kipili, Lake Tanganyika. Ad. 

Black occipital bar; 2 black annuli followed by 20 cross 
bars on the body. 
M. C. Z. 30403. Kasanga, Lake Tanganyika. Ad. 

Black occipital bar; 2 black annuli followed by a single 
cross bar, posteriorly a very few black blotches repre- 
sent rudiments of others. 
B. M. #3. Kasawa, Lake Tanganyika. Ad. 

Narrow, curved, black occipital bar; 2 black cross bars 
edged with lighter posteriorly on nape. 
B. M. #4. Nyasaland. Juv. 

Narrow, curved, black occipital bar; 3 black annuli 
followed by 7 black cross bars which become shorter 
posteriorly. 
B. M. #5. Nyasaland. Ad. 

Black occipital bar; 2 black annuli followed by 15 black 
cross bars on the body, a series of dark, light-centered, 
blotches on the flanks close to the ventrals, irregularly 
disposed between the dorsal bars. 

With regard to the British Museum specimens Mr. Parker adds, 
"The two groups appear distinct enough at first sight, particularly if 
only juveniles are considered. Numbers 5 and 6 are the only ones 
which might be considered as showing transition from one to the other. 
Reduction of the saddle-shaped cross-bars of number 6 on the dorsum 
and venter would leave some light centered blotches on the flanks 
comparable with those shown by number 5." 

Of the specimens in the Congo Museum at Tervueren, Dr. de Witte 
writes me that he has reexamined this material and that six snakes are 
referable to B. annulata. These are from Panga; Kobli; Uele; Umangi; 
Katanga and Poko respectively. To B. stormsi he would refer four 
snakes from Bosabangi; Albertville and Pweto. 

Variation. Midbody scale-rows 21-23; ventrals 200-209; anal 
single; subcaudals 71 in male, tail damaged in female; labials 7, the 
3rd and 4th entering the orbit ; temporals 1 +2 ; preocular 1 ; postoculars 
2; rostral about IJ^ times broader than deep. Schmidt has pointed 
out the unfortunate nature of the key suggested by Boulenger in 1904 
and based on the relative width and depth of the rostral. I might add 



266 bulletin: museum of comparative zoology 

that as the species lacks suboculars it is obvious that postocular is 
intended for subocular in the first part of the key while "two lower 
labials in contact with the lower subocular" in the second part is 
obviously a slip ; unfortunately some of these errors are repeated in his 
"List of the Snakes of West Africa." which was published in 1919 
(Proc. Zool. Soc. London, p. 293). 

Measurements. The specimen from Xgayu, Belgian Congo (A.M. 
X.H. 12329) listed by Schmidt is now in the Museum of Comparative 
Zoology. Curiously enough its total length is exactly that of the male 
from Kipili, liz. 1,385 (1,115+270) mm., though the tail of the 
Tanganyika snake is 5 mm. shorter than that of. the Congo reptile. 
Obviously there is nothing distinctive in the relative tail length of the 
two races. 

Hahitat and Habits. Kasanga is situated on delta flats in the centre 
of a sandy bay whose arms are formed by rocky promontories project- 
ing into Lake Tanganyika. On one of these the Germans built their 
important military base of Bismarckburg. Both promontories are 
protected by natural breakwaters of piled-up masses of rock. It is just 
off such rocks that the aquatic cobras are to be found. 

According to native reports, which my own experience confirmed in 
some points and contradicted in none, when the sun rises and strikes 
the rocks the cobras emerge from their retreats beneath them and 
bask for a short time on the tops of the rocks. Shortly afterwards, and 
I found none on the rocks an hour and a half after sun-up, they take 
to the water in search of fish. I was told that on a calm day one might 
see as many as ten in the course of a morning's fishing; we saw four in 
a little over three hours. The rocks slope precipitously beneath the 
water so that it is often ten feet deep within ten feet of the shore. I 
hired a boat and cruised very quietly along shore peering through the 
clear waters at the jumbled boulders, in and out from among which 
brilliantly colored small fish in great variety, darted or hovered. At 
last we saw a great head come out from beneath a rock followed by the 
handsomely barred neck and body of a large cobra which I estimated 
as about eight feet in length — native reports allege that they reach a 
length of ten feet which I do not think improbable. 

The natives were greatly excited and urged me to shoot it, so I fired 
first one and then the second barrel at the snake, which was between 
three and four feet below the surface; as I should have known, the 
bullets were deflected and did not penetrate the water; the only effect 
upon the snake Avas to cause it to retreat beneath the boulder. As I 
was waiting for its reappearance, a cry was raised from a watcher aft 



loveridge: African herpetology 267 

that a second snake was close by; glancing in the direction indicated, I 
saw a young three-foot cobra dart through the water with all the 
agility of an eel and disappear into a crevice among the submerged 
rocks. I concentrated on the first snake seen, however, and presently 
its head and a couple of feet of its body appeared. With my snake stick 
I pinned it down against a submerged rock, but the rock was slippery 
and with a few powerful convulsive jerks the reptile threw off the stick. 
Later the incident was repeated with another large cobra and I was 
forced to the conclusion that I was merely making myself ridiculous 
in attempting to hold down such powerful reptiles in their chosen 
element with nothing but a T-ended stick. I was discussing this after- 
wards with an Arab who told me that he had succeeded in catching 
one by a rather diabolical contrivance. He cut a stout V-shaped stick, 
then drove a long and strong nail into the apex of the fork; when the 
nail was firmly embedded in the wood, he filed away the head to a 
needle point. As he planted the stick upon the snake the nail pene- 
trated the backbone, disabling the reptile to some extent and prevent- 
ing its escape. I made a weapon according to this recipe but during 
the few hours left to me the opportunity to use it did not occur. 

That evening I visited the rocks at the north end of the bay as they 
would get the full benefit of the setting sun. The morning calm was 
gone, however, and the waves were pounding along the rocky shore 
with considerable violence. Everywhere I had been told that these 
snakes disliked a "rough sea" and leave the water when the wind 
rises. This is probably correct but at the same time it might be ob- 
served that when it is rough it is next to impossible to see down through 
the water to where the snakes might be ; even on a calm day the rippling 
of the water imparts an appearance of motion to every stick lying on 
the bottom so that these appear to be WTiggling! 

When the snakes come out of the lake in the evening they are said 
to bask rmdcr the rocks. This is probably correct for then they w^ould 
be sheltered from the boisterous, on-shore evening Avind which tends 
to cool the surface of the rocks, but they would receive heat radiating 
from the boulders above and below that had been exposed to the 
tropical sun all day. At the time of year — mid-May — when I was at 
Kasanga the evenings were distinctly cool. 

Presently one of the boatmen observed a snake lying far in betAveen 
tAA-o boulders but only just aboAC the reach of the breaking waACS. 
I landed and inspected the reptile Avhich mereh- moAcd still further in 
and concealed its head. Its tail was observed to be truncated and 
rotting and smelt quite offensively. The natiA-e explanation was that 



268 bulletin: museum of comparative zoology 

its tail had been bitten by one of its companions. Curiously enough 
only that morning I had shot a Nilotic Monitor Lizard whose tail was 
completely dead; the monitor was swimming in the lake and its tail 
was dried and withered but quite possibly might drop oflF in time. 

It was quite impossible to get at the snake in its present position, 
which was quite a yard in, so I retired another yard and fired a charge 
of dust-shot from a .22 calibre collecting gun. This caused the reptile 
to squirm, coil and uncoil. During these convolutions the tail came 
nearer and I slipped a leather noose over it, then started gently pulling 
till the tail was clear of the boulder, when I handed the noose-stick to 
Salimu with instructions to go on pulling slowly while I covered the 
snake with my T-ended stick until its head should appear. At that 
very moment the snake came away with unexpected suddenness, 
Salimu slipping on the wet boulder on which he had been standing, 
nearly fell, had a sudden vision of the cobra's head raised a foot from 
the rocks, dropped his noose-stick and fled incontinently. As a matter 
of fact there was nothing to worry about as I had pinned the snake 
about midbody with my T-stick, but it held its jaws wide open and 
Salimu thought that it was about to eject venom after the manner of 
its terrestrial relatives Naja and Sepedon. It distended its neck as it 
reared-up, but the spread was only half as broad as that of a cobra of 
the same size. 

The snake proved to be a five-foot female; had its tail been complete 
it would have been nearly six feet. Shifting my stick to its neck I re- 
noosed it behind the head and transferred it to a bag held by Salimu. 
As we turned to reembark in our boat the head and six inches of a 
cobra's body appeared above the waves about five feet offshore. "Its 
mate coming to look for it," said the boat boys with the native's gift 
of a theory for every circumstance. Though we waited some time it 
did not show itself again. 

Two days later we arrived at Kipili which lies about a third of the 
way up the east coast of the lake. The landing jetties, alongside which 
the lake steamers tie up, are at the sandy end of a big bay. On landing 
I made enquiries and was told that aquatic cobras were to be found 
only on the rocky coast of an island opposite the bay. I hired a dugout 
and set forth for the island but stopped on the way to examine the 
possibilities of a stretch of rocks which lay half a mile from the jetties. 
^^ e saw nothing at this spot and as there was an absence of suitable 
crevices in which the snakes could hide, I doubt if any but a stray 
cobra ever visits them. On clearing the headland we met the full force 
of a strong morning wind. Presently, soaked to the skin, and in momen- 



loveridge: African herpetology 269 

tary danger of being swamped by the big waves, we regretfully aban- 
doned the attempt. 

In the evening my dugout was placed on a motor launch and in this 
way crossed to the island which I was astonished to find, was fully four 
miles away! Here on a rocky islet only fifty feet in diameter, I dis- 
turbed a four and a half-foot cobra which slid down from a shelf beside 
me and into a crevice between the boulders where I shot it about mid- 
body as it was disappearing into the lapping water. Shifting the gun 
to my left hand I grabbed the snake's tail and only just in time for it 
would have disappeared in another second. For twenty minutes I 
held fast to its tail while the owner strained in the opposite direction 
and Salimu attempted to pry it free. When at last this was accom- 
plished, and despite the fact that its back had been broken by the dust- 
shot, no sooner did the poor beast's head come into view than it 
menaced us with open jaws after the manner of the Kasanga cobra, 
then it buried its fangs in its own body several times, holding on after 
each bite with the tenacity typical of the cobras. Quickly I slipped a 
noose over its head and transferred it to a bag where it was chloro- 
formed. As we continued to scour the rocks, peering into every crevice, 
a second snake was seen by one of the boat boys but made good its 
escape before I could reach the spot. 

Next day we reached Sumbwa to the north of Kipili; the wharf had 
been wrecked in a storm and so I was landed by boat upon the beauti- 
ful sandv beach. Naturallv no cobras could be found in such an en- 
vironment, but a Catholic missionary, who came off to the S.S. Liemba, 
told me that thev were abundant on the rocks at Karema, some seven 
miles away. Unfortunately we did not remain at Sumbwa long enough 
to permit of my visiting this place. 

Ujiji, five miles south of Kigoma, which is the western terminus of 
the Central Railway of Tanganyika, is also situated in a sandy bay 
and therefore without aquatic cobras though they are to be found at 
Bangwe, the northern headland of the l^ay. I went there in a dugout 
on May 28th, but the day was overcast and cold and though the whole 
morning was spent in searching for them, not a single snake was found. 
According to the local fishermen, these cobras are rarely seen out in the 
lake and are only to be encountered in the vicinity of rocks ; the same 
opinion was prevalent in each of the localities visited. It seems prob- 
able that at times they do so venture, otherwise they could scarcely 
have attained their present wide distribution in the lake. 

Defence. I came to the conclusion that, out of their element, they 
are not aggressive nor to be feared as much as the true cobras. Possibly 



270 bulletin: museum of comparative zoology 

their sight is not so good on land as under water; it would appear 
likely that it has undergone modifications to enable them to see clearly 
the fish on which they subsist. 

Venom. There is a widespread superstition prevalent among the 
fishermen that if a man is bitten in the water he should remain there 
until treatment in the shape of a water weed is brought to him. If 
he leaves the water he will rapidly succumb to the effects of the poison 
though if returned to the water for treatment hopes for his recovery 
may be entertained!! 

Parasites. Strangely enough, tliree ticks were found about the head 
of one of these aquatic snakes. 

Xaja melanoleuca Hallowell 

Naia haie var. melanoleuca Hallowell, 1857, Proc. Acad. Nat. Sci. Philad., p. 
61 : Gaboon, West Africa. 

1 (M. C. Z. 30405) Kitungulu, Urungu. 15. v. 30. 

1 (M. C. Z. 30406) Ukerewe Id., Lake Victoria, vi. 30. 

1 (M. C. Z. 30407) Entebbe, Lake Victoria. 27. vi. 30. 

1 (M. C. Z. 30408) Kampala, Uganda, vi. 30. 

1 (M. C. Z. 30409) Mabira Forest, Uganda. 1. vii. 30. 

Distribidion. Already recorded from Ukerewe Island by Sternfeld. 

Native name. Mufi (Kirungu). 

Variation. Midbody scale-rows 19; ventrals 206-210; anal entire; 
subcaudals 58-64; labials 7, the 3rd and 4th entering the orbit; post- 
oculars 3; anterior temporal 1. In all of which the series is normal. 

Measur emends. The largest snake, a male from Kitungulu, measures 
1,792 (1,486-^306) mm. 

Parasites. Nematodes {Kalicephalus sp.) and cestodes {Ophiotaenia 
iheileri) were in its stomach. The latter were reported on by Baer 
(1933, Revue Suisse Zool., p. 80) as from N. haje from "Uzungwe" a 
misreading of my label "Kitungulu, Urungu." 

Habitat. The Kitungulu snake, which was about to slough, was 
resting in a cavity in an enormous decayed tree trunk that was lying 
in a patch of swampy primary forest near the river bank. After we 
had been chopping at this log for nearly half-an-hour I left Salimu to 
finish it w^hile I went in search of a more remunerative one. Salimu 
subsequently stated that, shortly after my departure, this six-foot 
snake emerged from one end of the log and was making oft' when he 
stunned it with a blow, then bagged it. When he arrived in camp with 
it a couple of hours later it had fully recovered and was quite lively. 



LOVERIDGE: AFRICAN HERPETOLOGY 271 

The small Ukerewe Island specimen was taken just before my arrival 
by Pere A. Conrads who very kindly presented it to me for a record. 
Beneath a massive piece of rock near the Lake shore I found large 
quantities of cast skin and concluded that a cobra, or cobras, dwelt in 
the hole of which these were in the entrance. On the other hand when 
I related this to Pere Conrads he told me that the natives in that direc- 
tion reported losing cattle from time to time alleging that they were 
bitten by a big snake. This sounds more like the work of a mamba 
which Sternfeld has recorded from the island. 

I discovered the Entebbe snake beneath a log lying within ten feet 
of the lake shore and just as it was moving off, pinned it down with 
the handle of my frogging-net. Like all the other members of this 
series its stomach was empty. 

Naja nigricollis Reinhardt 

Naja nigricollis Reinhardt, 1843, Dansk. Vidensk. Selsk. Skrift, 10, p. 269, pi. 
iii, figs. 5 and 7: Guinea, West Africa. 

1 (M. C. Z. 30410) Kilimatinde, Ugogo. 26. xi. 29. 

2 (M. C. Z. 30411-2) Unyanganyi, Turu. 4-7. xii. 29. 

3 (M. C. Z. 30413-5) Mangasini, Usandawi. 12-14. xii. 29. 
1 (M. C. Z. 30416) Kitungulu, Urungu. 14. v. 30. 

Distribution. At Mpwapwa I examined a very large Black-necked 
Cobra of the grey black type which had been killed in a house in the 
town. The natives of Mwaya aver that the species occurs there, they 
are probably correct. Recorded by Sternfeld from Dar es Salaam, 
Victoria Nyanza etc. ' 

Native names. Ximdusu (Kikinga); niufi (Kirungu, but generic 
only). 

Variation. Midbody scale-rows 19; ventrals 196-205; anal entire; 
subcaudals 57-64; labials 6, only the 3rd entering the orbit; anterior 
temporals 2. In all of which the series is normal. 

Coloration. Above, black or smoke-grey; below, white, grey or 
black; all have a black tliroat and "neck" which in most of these 
specimens is followed by an area of pink more or less mottled with 
black. 

Measurements. Largest male (No. 30415) measures 1,441 (1,195-f- 
246) mm.; biggest female (No. 30410) measures 1,584 (l,312-|-272) 
mm. 

Breeding. Four of the snakes taken in December are very young; 
they range in length from 347 to 372 mm. 



272 bulletin: museum of comparative zoology 

Diet. Only one specimen, a Mangasini snake, held food, this was a 
frog {Rana adspcrsa). 

Parasites. Ticks were collected on the cobra from Kilimatinde, and 
nematodes {Kalieephahis sp.) in the Kitungulu snake. 

Ilabitaf. While at Kilimatinde I visited the old German fort which 
has been in ruins ever since the war. When I last saw it in 1926, it 
was so choked with vegetation that one could hardly get about in 
either courtyard or rooms. Recently the Church Missionary Society 
had the undergrowth cleared and commenced to reno\ate one of the 
buildings for use as a temporary hospital. On entering a cell-like room 
in this ruin I noticed a foot of cobra's tail protruding from a hole in 
the wall, creeping forward, I seized it, and at the same time placed my 
forked stick lightly on the portion nearest the hole. Naturally the 
snake pulled in an endeavor to withdraw itself completely into the 
wall and all I had to do was to keep up a gentle strain relaxing at inter- 
vals as one might play a fish. As was to be expected, after finding all 
its attempts to advance futile, the reptile reversed and started to come 
out backwards; several times by gentle pressure of the stick I checked 
its efforts at accelerated withdrawal. Two feet of snake were soon in 
sight, three, four and five followed and I was just beginning to wonder 
what sized snake I had to deal with when a second later its head came 
into view and was pinned down before it had a chance to spit; later I 
ascertained that the length was five and a half feet. Taking the cobra 
outside I put it in a bag from which it promptly escaped but instead of 
attacking, it made off but was recaught within a few yards. As I was 
returning it to the bag, it chewed on the latter so doggedly that I had 
much difficulty in getting it in; venom dripped from its fangs and 
trickled in beads down the bag and upon my fingers. Despite this 
abundance of venom the snake was much emaciated and I concluded 
that it had been aesti\ating through the exceptionally long dry season 
and had only recently emerged and set out in search of food. 

At Unyanganyi a young smoke grey cobra with an all-black collar 
extending to the back of its neck, came out from a hole at the base of 
a baobab tree just as a heavy shower on the 5th had ceased. Before it 
could get back I caught it and though it spat several times as I was 
securing it, the venom fell short. 

The Kitungulu snake crossed my path just after sunset and entered 
dry orchard forest five miles west of Kitungulu. 



loveridge: African herpetology 273 



Dendraspis angusticeps (Smith) 

Naia angusiiceps A. Smith, 1849, Illus. Zool. S. Africa, 3, pi. Ixx: Natal, South 
Africa. 

Aggressiveness. Curiously enough no mambas were seen during the 
whole trip. As doubts are sometimes thrown upon the aggressiveness 
of mambas it seems worth while recording the following accounts, both 
of which were given to me by gentlemen who were keenly interested 
in natural history. In neither case is the evidence absolutely convinc- 
ing that the snake was the aggressor though the probability is so in 
both, in the first story the snake may have assumed that it was itself 
about to be attacked. 

A small boy, child of a native squatter on the estate of my informant, 
Mr. Hardy, was herding his father's flock when his attention was 
attracted by the bleating of a goat. Hurrying in the direction of the 
sound he saw a mamba on, or wrapped about (?) the goat which it was 
biting; the snake then left the goat and came straight for the child and 
struck him with the result that the little goatherd died the same day. 

Mr. Fenwick of Miritini, near ^Mombasa, told me that the son of his 
neighbor, an Arab planter, was playing near the house to which he 
rushed back screaming and calling out that he had been attacked and 
bitten three times by a large snake. This child also died on the same 
day that he was bitten. Details of the story which I have since for- 
gotten pointed strongly to a mamba as being the species of snake 
concerned. 

VIPERIDAE 

Causus rhomb eatus (Lichtenstein) 

Sepedon rhotnbeatus Lichtenstein, 1823, Verz. Doubl. Mas. Berlin, p. 106: No 
localit}'. 

c? 9 (M. C. Z. 30417-8) Ilolo, Rungwe. 15. iii. 30. 
cf (M. C. Z. 30419) Kampala, Uganda, vi. 30. 

Distribution. Recorded from Bagamoyo and Bukoba by Tornier but 
these identifications may have been erroneous as they are not given by 
.Sternfeld who records it from Ukerewe Island. 

Native name. Kitumhi (Kinyakusa but very similar to their name for 
Viper a super cilia ris ) . 

Variation. Midbody scale-rows 19; ventrals 141-151; anal entire; 
subcaudals 22-29; labials 6. 



274 bulletin: museum of comparative zoology 

Measurements. The Kampala male measures 438 (400+38) mm., 
the Ilolo female 616 (550+66) mm. 

Diet. A toad (Bufo r. regiilaris) was present in the stomach of the 
Ilolo female. 

Venom. I was told that, just before the war, a European child had 
been bitten and died of snake-bite at the Rungwe Mission, Ilolo. 
That it was a Rhombic Night xVdder seems probable though Sternfeld 
has recorded Cau.nis dcfiUppii as occurring at Tukuyu which is only 
ten miles from Ilolo. 

Causus resimus (Peters) 

Heterophis resimus Peters, 1862, Monatsb. Akad. Wiss. Berlin, p. 277, pi. 
— , fig. 4: Gebel Ghule, Senaar, Sudan. 

9 (M. C. Z. 30420) Ukerewe Id., Lake Victoria, vi. 30. 

Distribution. Recorded from Dar es Salaam by Sternfeld and from 
Lake Tanganyika by Boulenger. 

Variation. Midbody scale-rows 21; ventrals 143; anal single, sub- 
caudals 18; labials 6. 

Coloration. The uniformly plumbeus appearance of the preserved 
snake gives a poor idea of the wonderfully vivid, yet velvety, green 
color of the living night adder. It is one of the most beautiful of 
East African reptiles. 

Measurements. Total length 424 (390+34) mm. 

Causus defilippii (Jan) 

Heterodon defilippii Jan, 1862, Arch. Zool. Anat. Phys., 2, p. 225: Africa. 

1 (M. C. Z. 30421) Mpwapwa, Ugogo. 23. xi. 29. 
1 (M. C. Z. 30422) Kitungulu, Urungu. 16. v. 30. 

Distribution. According to Sternfeld, Defilippi's Night /Vdder also 
occurs at Bagamoyo, Dar es Salaam, Tukuyu and in Ugogo. 

Variation. Midbody scale-rows 17; ventrals 114-116; anal single; 
subcaudals 14-15; labials 5-6; the former resulting from a fusion of the 
3rd and 4th making the 3rd, instead of the 5th, labial the largest; the 
frontal is equal to, or longer, than its distance from the end of the 
snout. 

Breeding. Both specimens are very young being only 234 (212+22) 
and 155 (142+13) mm. respectively. 

Habitat. I found the Mpwapwa snake under the same rotting tree 



loveridge: African herpetology 275 

stump in sandy soil where I took Rhinocalamus dimidiatus. The 
Kitungulu reptile was driven out by the fire which we set to a pile of 
refuse in a native garden. 

ViPERA superciliaris Peters 

Vipera superciliaris Peters, 1854, Monatsb. Akad. Wiss. Berlin, p. 625: Cape 
Delgado, Mozambique; 1882, Reise nach Mossamb., 3, p. 144, pi. xxi; 
Pfeffer, (1892) 1893, Jahr. Hamb. Wiss. Anst., 10, p. 89: Quilimane, 
Mozambique; Boulenger, 1915, Proc. Zool. Sec. London, p. 638: "German 
East Africa at Cape Delgado"; Cott, 1928, Proc. Zool. Soc. London, p. 
934: Charre and Fambani, Mozambique. 

2 (M. C. Z. 30423^) Mwaya, Lake Nyasa. 1-8. iii. 30. 

Distribution. These are presumably the sixth and seventh recorded 
examples of the rare Yellow-browed Viper. They constitute the first 
record for Tanganyika Territory for Boulenger was in error in 1915 
when he transferred Cape Delgado from Mozambique to German East 
Africa. 

Native name. Katumbi (Kinyakusa, very similar to that applied to 
Cavsus rhomheatus) . 

Variation. Both agree with the type in the number of midbody 
scale-rows 27; but they have 140 instead of 142 ventrals; subcaudals 
40 as in the type with which they agree in other respects. It would 
appear as if the range of variation is small in this ancient, isolated, and 
most southerly member of its genus in Africa. 

Coloration in life. These examples were richer and even more hand- 
some than the specimen figured by Peters. 

Measurements. Both males, the larger 488 (425+63) mm., the 
smaller 283 (245+38) mm., both being surpassed by the type which 
was 570 mm. The Banyakusa at Mwaya probably confuse this 
snake with the Puff Adder for they told me that it attains the size of 
a man's arm and was not uncommon. 

Diet. The larger held a rat in its stomach and the smaller a frog 
(Phrynobatrachus acridoides). All that we know of the habits of this 
snake in life is given by Cott, who writes: "Several specimens of a 
very rare snake, Vipera superciliaris, which is known locally as "Tan- 
daruma," were taken. This species, which it appears has only been 
found once previously, is for some reason extremely difficult to keep 
in captivity; specimens from Charre and Fambani invariably refused 
food, and all died within a week or two of being captured." 

I would suggest that the early deaths imply that the snakes had 



276 bulletin: museum of comparative zoology 

been internally injured by their captors if these were natives, for a 
blow on the spine or congestion caused by being carried in a split stick 
would render a snake indisposed to feed and cause its death apart from 
malnutrition. On the other hand, though their refusal of food could 
not cause death in so short a time, it may be that these vipers share a 
disinclination to feed in captivity with their European ally Vipera b. 
bervs. 

BiTis ARiETANS (Mcrrcm) 

Vipera arietans Merrem, 1880, Vers. Sjst. Amphib., p. 152: Cape of Good 
Hope. 

1 (M. C. Z. 30425) Mpwapwa, Ugogo. 22. xi. 29. 

3 (M. C. Z. 30426) Mangasini, Usandawi. 13. xii. 29. 

1 (M. C. Z. 30427) Igale, Poroto Mtns. 30. iv. 30. 

3 (M. C. Z. 30428) Ukerewe Id., Lake Victoria. 14. vi. 30. 

1 (M. C. Z. 30429) Kampala, Uganda, vi. 30. 

Distribution. Recorded by Sternfeld from Kitopeni; Dar es Salaam; 
Mpwapwa; Lake Tanganyika; Bukoba and Ukerewe Island; by 
Boulenger from Lake Nyasa. 

Measumncnts. The largest male was from Ukerewe Island and 
measured 1,151 (1,020+131) mm. or 45^ inches, surpassing my 
largest (Kilosa) record by two inches. While such big snakes are 
extremely rare at Kilosa they appear to be of normal occurrence on 
Ukerewe for this was only one of three large Puff Adders — all males 
over 423/^ inches — brought in together slung on a pole. Several others 
of similar dimensions were also offered for sale but not purchased. 
The midbody circumference of the largest male was 190 mm. 

Diet. Stomach contents consisted of: — (1) a rat {Rattus r. kijabus) 
at Mangasini; (2) a rat (Bhabdovu/s p. dimidiaius) at Igale; (3) a 
rodent in one of the Ukerewe Island snakes. 

Parasites. Numerous nematodes (Ophidascaris sp. &: Thuhuneasp.), 
some of extraordinary length, were found in the Ukerewe specimens. 

BiTis GABONiCA (Dumeril & Bibron) 

Echidna gabonica Dumeril & Bibron, 1854, Erpet. Gen., 7, p. 1428, pi. Ixxxb: 
Gaboon, West Africa. 

d" (M. C. Z. 30430) Mbuyu near Kampala, Uganda, vi. 30. 

Variation. Midbody scale-rows 34; ventrals 128; anal single; sub- 
caudals 30; labials 14. 



loveridge: African herpetology 



277 



Measurements. Total length 355 (320+35) mm. 
Diet. A mouse {Leggada sj).) was in its stomach. 



Atheris barbouri Loveridge 

Athens harbouri Loveridge, 1930, Proc. N. Eng. Zool. Club, 11, p. 107: Dabaga, 
Tanganyika Territory. 

3 (M. C. Z. 29055-7) Dabaga, Uzungwe Mtns. 1. i. 30. 

6 (M. C. Z. 30431-5) Madehani, Ukinga Mtns. 13-18. ii. 30. 

Distribution. In addition to these locaHties I was shown a specimen 
taken near ^lufindi at the southern end of the Uzungwe Mountains 
and a missionary at Tandala in the Ukinga Mountains described one 
of these tree vipers which she had killed in her garden the week 
previous to my arrival, but a species which she had no recollection of 
having seen before during her many years of residence there. 

Xative names. Moma (Kikinga); mboma (Kirungu). Both these 
forms are used in Kiswahili for the Puff Adder, in all probability they 
are applicable to any viper. 

Variation. Since the publication of the original description which 
was written in the field and based on the three Dabaga snakes the 
additional data available from a study of the Madehani series permit 
of a slight extension of the range of variation. As all five species of 
the genus are represented in the collection of the Museum of Compara- 
tive Zoology I have tabulated the available data as regards the more 
important scale-counts most of which have been considerably in- 
creased since the publication of the Catalogue of Snakes. Schmidt has 
suggested that Boettger's A. laeviceps from Boma, Belgian Congo 
might be revived as a race of squamigera but for the purposes of this 
tabulation they are treated as one species. 







Midbody 










Scales 


Scales 






scale- 






Sub- 




round 


across 




Species 


rows 


Ventrals 


Anal 


caudals 


Labials 


orbit 


occiput 


A. 


barbouri 


19-23 


114-128 




14-22 


5-6  


8-13 


8-9 


A. 


ceratophorus 


21-25 


142-152 




54^56 


9-11 


16-17 


8-11 


A. 


nitschei 


25-32 


141-162 




35-52 


9-12 


12-15 


8-10 


A. 


chloroechis 


25-36 


154-165 




48-62 


9-13 


15-20 


9-14 


A. 


squamigera 


15-25 


153-173 




45-65 


7-12 


10-18 


6-9 



From the above data I conclude that squamigera is the most primi- 
tive species, it is also the most widely distributed, and that the others 
are offshoots of which A. barbouri is an end form occurring far to the 



278 bulletin: museum of comparative zoology 

southwest of any other members of the genus. Perhaps it might be as 
well to remark that in reality the members of the genus are much 
more distinct than the tabulated data might lead one to suppose. 

Coloration. See Habitat. 

Measurements. The largest of four males measures 352 (315+37) 
mm.; the type still remains the largest of the five females being 369 
(335+34) mm. 

Sex. Each specimen was carefully sexed and no overlapping of 
scale-counts occurs in this small series. 

d" ventrals 114-118; subcaudals 20-22. 
9 " 119-128; " 14-19. 

Breeding. A female taken at Madehani on 13. ii. 30 held ten eggs 
measuring 10 x 6 mm. Other females brought in on the 14th and 18th 
respectively also each held ten \ ery small eggs, while a fourth female 
held none on the 18th. One snake was very young and only measured 
1G4 (148+16) mm. when found in the road on the 19th. 

Diet. A large earthworm was taken from the stomach of the young 
snake to which I have just referred. 

Habitat. A woman had dug up a pair when hoeing the ground for 
planting, I imagine that they were concealed among sods such as 
litter the gardens. She brought them alive. The male had rather in- 
distinct markings abo\'e and was uniformly olive below. The female 
had a chain of rhombs down the back, was oli^•e below but chequered 
with black posteriorly. 



Atractaspis irregularis (Reinhardt) 

Elaps irregiilaris Reinhardt, 1843, Dansk, Vidensk. Selsk. Skrift., 10, p. 264, 
pi. iii, figs. 1-3: Gaboon, West Africa. 

1 (M. C. Z. 30275) Entebbe, Lake Victoria. 28. vi. 30. 

Distrihution. Recorded by Boulenger from I'ganda and by Stern- 
feld from the Zanzibar coast. 

Affinities. In view of the fact that we now know so many African 
Atractaspis have extensive ranges across Africa from coast to coast 
as is the case with irregularis, bibronii and aterrima, and also that they 
exhibit a greater degree of variation than was originally supposed, it 
seems very possible that A. bipostocularis Boidenger from Mt. Kenya 
may ultimately have to be united with irregularis. There are two 
examples of other species in the collection of the Museum of Compara- 



loveridge: African herpetology 279 

tive Zoology which possess a single postoeular on one side of the head 
and a pair on the other! 

Variation. This Entebbe specimen agrees perfectly with the de- 
scription of irregtdaris given in the Catalogue of Snakes except that, 
instead of 25-27, it possesses 23 midbody scale-rows. Ventrals 229; 
subcaudals 24; labials 5, the 3rd and 4th entering the eye; postoeular 1. 

When passing through Nairobi, I took the opportunitj' of reexamin- 
ing the Nairobi Museum specimen No. I 92 without locality which 
agrees with the present specimen in possessing 23 midbody scale-rows. 
It is, however, a typical irregularis in every other respect except that 
the 4th lower labial instead of the 3rd is the largest. The 4th lower 
labial being largest is, together with the two postoculars, the key 
character distinguishing bipostocularis referred to above. 

Measurements. The Entebbe snake is very young with umbilical 
scutes still unhealed, it measures 247 (230+17) mm. 

Diet. Stomach contents consisted of a Lcptotyjjhlops conjuncta 
though the identification is an assumption for the head of the prey is 
completely digested away, what remains measures 127 mm. 

Atractaspis conradsi Sternfeld 

Atradaspis conradsi Sternfeld, 1908, Sitzb. Ges. Naturf. Freunde Berlin, p. 94: 
Ukerewe Island, Tanganyika Territory; Roux, 1910, Ann. Zool. Suisse, 
p. 99: Bukoba, Tanganyika Territory; Sternfeld, 1912, Wiss. Ergebn. 
Deutsch-Zentral-Afrika-Exped., 4, p. 278: Lake Kivu, Belgian Ruanda; 
Boulenger, 1915, Proc. Zool. Soc. London, p. 640: Entebbe, Uganda. 

Distribution. Ukerewe Island was visited with the express purpose 
of securing a series of these burrowing vipers ; unfortunately, however, 
the rainy season w^as long past and the countryside already somewhat 
parched so that I failed to obtain any specimens of conradsi during my 
ten days' stay. Through the courtesy of Pere Conrads, the original 
discoverer of this snake, I w^as permitted to examine his recent col- 
lections among which I found a young Atractaspis of considerable in- 
terest which is undoubtedly referable to bibronii, of which rostrata is a 
synonym. 

The interest lies in the fact that it shows that bibronii occurs in 
three of the localities from which conradsi has been reported, viz. 
ITvcrewe Island, Bukoba and Entebbe. 

Affinities. The only character which differentiates A. conradsi from 
A. bibronii is that the anal and subcaudals of the former are divided 
or paired while in bibronii they are single. In view of the variation 



280 bulletin: museum of comparative zoology 

shown by West African Atracfaspis it remains to be seen whether it is 
of specific importance in the present instance. 

The original description of conradsi only occupies four lines and a 
translation reads as follows: — "Near A. irregularis from which it is 
distinguished by a somewhat pointed snout and by the 23 scale-rows. 
Color dark blackish-brown. Length 50 cm. Tail 2.8 cm. 1 Ex. V = 
257. Sq =23. Sc =23. Ukerewe Id., D. O. A. Conrads." When re- 
cording the Kivu specimen in 1912, he added that the sutures betw^een 
the prefrontals and nasals are of practically the same length; sym- 
phisial is separated from the anterior chin-shields; 3rd lower labial 
enormous; anal and subcaudals up to the last, divided. Roux also 
gives the data of his snake. 

Variation. In the young hibronii from Ukerewe Island in Pere 
Conrads' collection, the snout is pointed; there is a single postocular in 
contact with a large temporal; 1st lower labial on the left side reaches 
the median line of the throat, its fellow on the right does not but is well- 
separated, if one assumes that the aljnormality is of the right side, 
then, if normal, the labials would be in contact; midbody scale-rows 
25; ventrals 223, anal entire, subcaudals single 21; upper labials 5 of 
which the 4th is much the largest. Thus though the snake is un- 
doubtedly hibronii it will be observed that only in the matter of the 
anal and subcaudals is it distinct from conradsi for it is probable that 
when more material is a^•ailable, conradsi will be found to have a mid- 
body scale count of from 23-27 like the allied species. 

Atractaspis bibronii Smith 

Airadaspis bibronii A. Smith, 1849, Illus. Zool. S. Africa, 3, pi. Ixxi: Eastern 

districts of Cape Colonj-, South Africa; Schmidt, 1923, Bull. Am. Mus. 

Nat. Hist., 49, p. 138: Garamba, Belgian Congo. 
Atractaspis rostrata Giinther, 1868, Ann. Mag. Nat. Hist. (4). 1, p. 429, pi. 

xix, fig. 1: Zanzibar; Barbour & Loveridge (part), 1928, Mem. Mus. 

Comp. Zool., 50, p. 137; Dar es Salaam and localities in the Uluguru 

Mountains, Tanganyika Territory. 

Affinities. The Museum of Comparative Zoology having very 
recently obtained examples of bibron.ii from both southwest and south- 
east Africa, I have taken the opportunity of checking the conclusions 
of Werner, which were followed by Schmidt though the latter had no 
southern material for study. I fully concur with these authors and 
fail to find any characters whereby one may distinguish rostrata (East 
Africa) from the older hibronii (South Africa). This brings into line 



loveridge: African herpetology 281 

Sternfeld's records of both "species" which he reports from various 
localities in Tanganyika Territory. Roux has recorded rostrafa from 
Bukoba and Boulenger from Lake Nj'asa and Uganda. See remarks 
under Atradaspis conradsi. 

Atractaspis aterrima Giinther 

Atradaspis aterrima Giinther, 1863, Ann. Mag. Nat. Hist. (3) 12, p. 363: West 
Africa; Boulenger, 1915, Proc. Zool. Soc. London, p. 640: Uganda and 
West Africa, from the Gold Coast to the Niger. 

Atractaspis rostrata (part) Barbour & Loveridge, (nee. Giinther), 1928, Mem. 
Mus. Comp. Zool., 50, p. 137; part Nyange series only. 

d' (M. C. Z. 23466) Nyange, Uluguru Mtns., 11. x. 26. 

Distrihiition. This record involves a considerable extension of range 
to the eastward for hitherto Uganda was the most easterly record. 

Variation. This specimen is one of a series of three snakes from 
Nyange which were referred by Barbour & Loveridge to rostrata 
( =bibronii). While the other two are referable to hihronii, of which 
rostrata is now considered a synonym, this snake must be referred to 
aterrima for it possesses 27 ventral scutes more than any of the others. 

In connection with checking the alleged differences between hihronii 
and rostrata my attention was attracted to this snake by its rounded 
snout which definitely separated it from all the sharp-snouted hihronii 
in the collection. On comparing it with the description of aterrima I 
found it to agree in all respects except that the midbody scale-rows 
were 23 instead of 19-21 as recognised for aterrima. I feel confident 
that the range for aterrima should be increased to 19-23. 

This specimen has 276 ventrals, an undivided anal and 25 unpaired 
subcaudals. This correction makes it necessary for the range of ven- 
trals in hihronii to revert to their former range of 221-260, not 276. 

Measurevients. Total length 559 (525 -f- 34) mm. 



Atractaspis microlepidota Giinther 

Atractaspis microlepidota Giinther, 1866, Ann. Mag. Nat. Hist. (3), 18, p. 29, 

pi. vii, fig. 3: Type locality unknown. "Probably West Africa" errore; 

Loveridge, 1916, Journ. East Afr. & Uganda Nat. Hist. Soc, p. 87: No 

locality, probabh^ Kenya. 
Atractaspis phillipsi Barbour, 1913, Proc. Biol. Soc. Wash., p. 148: Singa, 

Senaar province, Sudan.; Boulenger, 1915, Proc. Zool. Soc. London, p, 

658: Key to species. 



282 bulletin: museum of comparative zoology 

Atradaspis magretti Scortecci, 1929 (1928), Atti. Soc. Italia. Sci. Nat. Mus. 
Civ. Milano, 67, p. 308, fig. 6: Mandafena & Monte DongoUa, Erythraea. 

1 (M. C. Z. 29999) "Kenya Colony." 

Distribution. A. microlepidota has been recorded by Boulenger from 
Lake Tanganyika and by Sternfeld from Lamu, Kenya Colony. 

Variation. The specimen listed above, whose full data is given in 
the 1916 citation, was originally I 93 in the collection of the East 
Africa and Uganda Natural History Society by whom it was given to 
me to bring back to the Museum of Comparative Zoology. 

I have carefully compared it with the figure of Giinther's type and 
find them in full agreement except that it possesses two postoculars 
on one side of the head, but only one, like the type, on the other. 

I have also compared the type of .4. phiUipsi with the figure and 
with this specimen and find that phillipsi only differs in that the 4th 
labial alone enters the orbit, instead of the 3rd and 4th in microlepidota. 
However the two sides of the head in pJiillipsi are not entirely alike 
for on the left side of the head the 3rd labial very nearly enters being 
separated by a space and not a scale. Except for some very trifling 
variation in the relative proportion of some of the head scales the three 
snakes are in full agreement. 

A. phiUipsi and A. magretti undoubtedly came to be described on 
account of Boulenger's erroneous key in the Catalogue of Snakes, 3, p. 
512 and repeated in the 1915 citation given above. These keys ignore 
the scale, cut off from the upper part of the 5th labial which Barbour 
rightly calls "a single large anterior temporal" for in other species the 
analogous scale is so-called by Boulenger himself, in some specimens 
it is semi-posterior to the postocular, in others almost below it; it is 
clearly shown in Giinther's figure of the type. In ignoring it Boulenger 
states "temporals small" as a major division in his key as opposed to 
"Postocular in contact with a large temporal." 

GEKKOXIDAE 

Paragonatodes quattuorseriatus (Sternfeld) 

Gonatodes quattuorseriatus Sternfeld, 1912, Wiss. Ergebn. Deutsch-Zentral- 
Afrilia-Exped., 4, p. 202, pi. vi, fig. 1: Kissenje; Uvira; Lake Kivu, etc., 
Belgian Ruanda. 

cf (M. C. Z. 3043G) Mpwapwa, Ugogo. 22. xi. 29. 

Distribution. The Mt. Kenya specimen of P. africanus has since 
been referred to quattuorseriatus by Nieden. One would have expected 



loveridge: African herpetology 283 

the Mpwapwa lizard to be referable to africanus which occurs on the 
Usambara Mountains (its type locality), the Uluguru Mountains, Mt. 
Meru, and Kilimanjaro (loc. incert) ; this is not the case, however. 

Variation. The Mpwapwa gecko has 7 upper labials, 5 lower labials 
and 7 preanal pores. The two species are distinguished as follows: — 

5-6 upper; 5-6 lower labials; 7-8 preanal pores .... quattuorseriatus 

6-7 upper; 7-8 lower labials; 8-12 preanal pores . . . africanus 
One might be tempted to suppose that they are not specifically dis- 
tinct were it not for the fact that Sternfeld based his description on 
nine cotypes. Of africanus we have over a score of topotypes in the 
Museum of Comparative Zoology and none of these intergrade. The 
Mpwapwa gecko has been compared with one of Sternfeld's cotypes 
also in this collection. 

Coloration. The coloration of the Mpwapwa gecko is identical with 
that of Usambara africanus; according to Sternfeld the coloration of 
quattuorseriatus is the same as that of africanus "but brighter." 

Measuremenis. Total length 60 (36+24) mm. 

Habitat. Taken in sandy debris among the rotting roots of a fallen 
tree lying fifty feet from a small stream but in dry and dusty ground. 
It is a rain-forest form surviving in a locality which is undergoing 
desiccation. 

Hemidactylus mabouia (Moreau de Jonnes) 

Gecko mabouia Moreau de Jonnes, 1818, Bull. Soc. Philom. Paris, p. 138: 
Antilles and adjacent mainland. 

9 (M. C. Z. 30437) Bagamoyo. 11. xi. 29. 

9 (M. C. Z. 30438) Mangasini, Usandawi. 13. xii. 29. 

d" (M. C. Z. 30439) Mwaya, Lake Nyasa. 1. iii. 30. 

Distribution. Also seen on Mombasa Island and at Changamwe, 
Tanga, Dar es Salaam and Kilimatinde. 

Native names. Zirambi (Kisandawi); kanakipili (Kinyakusa). 

Variation. Male with 50 preanal pores; 9-12 rows of conical dorsal 
tubercles; 7-8 subdigital lamellae under the median digit. 

Coloration. Grey on stem of cultivated banana at Changamwe; 
white on whitewashed wall of Hotel Africa, Dar es Salaam. 

Enemies. Two were recovered from the stomachs of Spotted Wood 
Snakes (Philothamnus s. semivariegatus) at Bagamoj'o. 

Habitat. Numerous on walls of a deserted Arab building at Baga- 
moyo; on rocks in dry river bed at Kilimatinde; on rocks on kopjes at 
Mangasini, and very abundant on the trunks of trees forming the main 
avenue at Mwaya, a habitat which they share with Agama atricollis. 



284 bulletin: museum of comparative zoology 

Hemidactylus persimilis Barbour & Loveridge 

Hemidadylus persimilis Barbour & Loveridge, 1928, Mem. Mus. Comp. Zool., 
50, p. 140, pi. iv, figs. 1 and 3: Dar es Salaam, Tanganyika Territory. 

cf (M. C. Z. 30440) Miritini, Kenya Colony. 30. x. 29. 
5 9 9 & eggs (M. C. Z. 30441-5) Bagamoyo. 11. xi. 29. 

Young (M. C. Z. 30446) Morogoro, Ukami. 20. xi. 29. 

Variation. Male with 34 preanal pores; 12-17 rows of conical dorsal 
tubercles; 5 subdigital lamellae under the median digit. 

Breeding. Each of the Bagamoyo females clearly showed a pair of 
ovarian eggs tlu-ough the abdominal skin, at the same time and place 
fifteen pairs of eggs were taken, the units of each pair, unlike those of 
H. mabouia, were separate in each instance ; both geckos and eggs were 
in the mcuti thatch of collapsed huts in native rice fields. 

Enemies. One gecko was recovered from the stomach of a Green 
Snake {CJdorophis hoplogaster) at Bagamoyo. 

Habitat. The Miritini male was taken under a piece of burnt bark 
and was black in consequence ; the young Morogoro gecko was on the 
stem of a banana close to some huts. 

Hemidactylus tropidolepis squamulatus Tornier 

Hemidadylus squamulatus Tornier, 1897, Die Kriechthierc D-O-A, p. 10: 

Kakoma, Ugundu, Tanganyika Territory. 
Hemidactylus tropidolepis Barbour & Loveridge (non Mocquard), 1928, Mem. 

Mus. Comp. Zool., 50, p. 142: Near Kilindini, Kenya Colony. 

cT (M. C. Z. 20447) Near Kilindini, Kenya Colony. 29. x. 29. 

Affijiitics. Recently Parker (1932, Proc. Zool. Soc. London, p. 342) 
with fresh Somaliland material has raised the question of the distinct- 
ness of squamulatus, long considered a synonym of tropidolepis and 
considers the latter distinct on the basis of the fewer preanal pores 7-8 
as against 13-19 in squamulatus. 

Variation. Male with 15 preanal pores; 3 pairs of chin-shields. 

Measurements. Total length 65 (36+29) mm. 

Habitat. Obtained at the same spot as the two examples collected 
in 1926, viz. on the mainland opposite Kilindini harbour. Apparently 
the species is decidedly rare for I spent some hours in unavailing 
search both before and after this gecko was found between some dead 
palm fronds piled upon a stump in a native garden. 



loveridge: African herpetology 285 

Hemidactylus werneri w erneri Tornier 

Hemidactylus nrrneri Tornier, 1897, Arch. Naturg., 63, p. 63: Dalalani, Tan- 
ganyika Territory. 

Hemidactylus werneri werneri Loveridge, 1929, U. S. Nat. Mus. Bull. No. 151, 
p. 44: Itende (not Hende), Dodoma, Tanganyika Territory. 

d^ (M. C. Z. 30464) Mpwapwa, Ugogo. 23. xi. 29. 
9 (M. C. Z. 30465) Masiliwa, Turn. 10. xii. 29. 
9 (M. C. Z. 30466) Maji Malulu, Usandawi. 10. xii. 29. 
9 (M. C. Z. 30467) Mangasini, Usandawi. 13. xii. 29. 

Variation. Upper labials 6-7; lower labials 5-7; lamellae under 
median digit 4-7 pairs. 

Measurements. None exceed in size specimens previously recorded. 

Habitat. The Mpwapwa gecko was taken under a rotting tree stump 
in sandy soil forming the bank of a dry watercourse. The Masiliwa 
specimen from beneath the bark of a fallen tree at 9 a.m. The one 
from Maji ^Nlalulu at dusk, close to a hole into which it attempted to 
retreat. At Mangasini, just after my tent had been pitched, I cap- 
tured a half-grown gecko which was running up the inside of my tent; 
presumably it had been disturbed during the hoeing-over of the camp 
site, though it is feasible to suppose that it may have been brought in 
the tent from the last camp at Maji Malulu. I mention this in view of 
the fact that no other werneri were taken during our week's stay at 
Mangasini. 

Hemidactylus werneri alluaudi Angel 

Hemidactylus alluaudi Angel, 1923, Bull. Mus. d'Hist. Nat. Paris, p. 490: Bura, 

Kenya Colony. 
Hemidactylus werneri alluaudi Loveridge, 1929, U. S. Nat. Mus. Bull. No. 151, 

p. 46. 

As I was passing through Paris, through the courtesy of Mons. F. 
Angel, I was afforded the opportunity of examining the type of this 
gecko which confirmed my views as to its very close affinity with 
H. werneri Tornier. 

The holotype is a female, distended with ova, possessing 62 midbody 
scale-rows and 16 rows of enlarged and strongly keeled scales across 
the back though these are not more strongly keeled than in werneri; 
upper labials 6 (it is a matter of opinion whether they can be con- 
sidered 7 or 8 as the posterior ones are scarcely differentiated) ; 5 lower 
labials; 5 pairs of lamellae under the median digit, the distal pair un- 
divided. The coloration is the same as in werneri but is much faded. 



286 bulletin: museum of comparative zoology 

There remains therefore, the one character of the mental separating 
the chin-shields which differentiates this form from typical iverneri 
where the chin shields are in contact. 



Hemidactylus brookii Gray 

Hemidadylus brookii Gray, 1844, Zool. Erebus and Terror, pi. xv, fig. 2: 
"Australia; Borneo." 

3 (M. C. Z. 30448-50) Saranda, Ugogo. 18. xi. 29. 
20 (M. C. Z. 30451-60) Ukerewe Id., Lake Victoria. 10-19. vi. 30. 
3 (M. C. Z. 30461-3) Kampala, Uganda, vi. 30. 

Variation. Males with 30-34 preanal pores, average 31. 

Measurements. The largest of 14 males (No. 30460) measures 125 
(65+60) mm. though the tail is reproduced; the largest of 12 females 
is 122 (62+60) mm. One Saranda gecko is very young being only 58 
(28+30) mm. 

Habitat. At Saranda, where Pachydactylus houlengeri was found on 
the walls of the houses, these geckos were taken beneath the bark of 
fallen trees and in a hole in a tree-trunk. At Ukerewe Island, in the 
absence of the house geckos H. mabouia and P. boiilengeri, this species 
was abundant on the walls of the mission outbuildings and my first 
specimen was taken at 10 p.m. on the day of arrival on a post of the 
verandah at the Mission. At Morogoro, where both H. mabouia and 
H. brookii occur the latter lives in the bush or under rubbish while 
H. mabouia occupies houses and large trees. Lang has already pointed 
out that brookii is the common house gecko in the Congo. 

Lygodactylus capensis capensis (Smith) 

Hemidactylus capensis A. Smith, 1849, Illus. Zool. S. Africa, 3, pi. Ixxv, fig. 3: 
Kaffirland and districts north of Cape Colony. 

4 (M. C. Z. 30487-90) Masiliwa, Turu. 10. xii. 29. 

Variation. After comparison with South African capensis I fail to 
find any differences sufficiently marked to justify one in differentiating 
East African specimens. Males with 7 preanal pores; upper labials 
6-8; lower labials 6-7; 4 pairs of lamellae under the longest digit. 

Coloration. The grey color and markings of these geckos so closely 
matched the lichen-covered bark on which they were found as to 
render their detection difficult. 

Measurements. The larger of two males measures 65 (32+33) mm. ; 



loveridge: African herpetology 287 

the larger of the females 61 (29+32) mm. The smaller is very young 
being 3G (20+16) mm. 
Habitat. Taken on the trunks of trees growing in open woodland. 

Lygodactylus stevensoni Hewitt 

Lygodadylus stevensoni Hewitt, 1926, Ann. Natal Mus., 5, p. 445, pi. xxv, 
figs. 3-4: Khami Ruins, S. Rhodesia. 

d" {M. C. Z. 30491) Nyamkolo, N. Rhodesia. 9. v. 30. 

Distrihition. Geckos, which appeared to be specifically identical 
with this species, were seen on the Stevenson Road near Ikombo, N. 
Rhodesia and in the vicinity of Kitungulu, Urungu. 

Variation. Males with 8 preanal pores; upper labials 8; lower 
labials 6; 4 pairs of lamellae under the longest digit. 

This lizard agrees with Hewitt's description of the three cotypes in 
its more pointed snout, nostril and first labial arrangement and other 
characters. I fail to observe the faint indications of caudal segmenta- 
tion of which he speaks. 

Measuremenis. Total length 65 (30+35) mm. 

Enemies. It was presumably this species which was recovered from 
the stomach of a snake (Amplorhinus nototaenia) at Kitungulu. 

Habitat. The Ikombo gecko, which made its escape, was under a 
log in dry maiombo orchard-bush. 

Lygodactylus grotei Sternfeld 

Lygodadylus grotei Sternfeld, 1911, Sitzber. Ges. Naturf. Freunde Berlin, 
p. 245: Mikindani and Makonde Highlands, Tanganyika Territory. 

3 and eggs (M. C. Z. 30492-5) Bagamoyo. 11. xi. 29. 

9 (M. C. Z. 30496) Morogoro, Ukami. 20. xi. 29. 
9 (M. C. Z. 30497) Shinyanga, Usukuma. 4. vi. 30. 

Affinities. While reexamining part of the type series of L. capensis 
mossambica Loveridge I fail to find scale characters which will dis- 
tinguish it from grotei. L. c. mossambica appears to be an intermediate 
between capensis and grotei having the coloring of the former and the 
subcaudal arrangement of transversely enlarged scales which charac- 
terises the latter. I hesitate to unite them for when I collected the 
type series of fifty mossambica I had come fresh from collecting more 
than a score of grotei and the Mozambique specimens struck me as 



288 bulletin: museum of comparative zoology 

being different. Some fresh material from Lumbo should settle the 
point. 

Variation. Upper labials 7-9; lower labials 5-8; 4 pairs of lamellae 
under the longest digit; preanal pores in male 5. 

Mcasuremcnis. Single male 59 (27+32) mm., largest female 62 (33 
+29) mm. 

Breeding. Two eggs measuring 6x5 mm. were found under palm 
fronds at Bagamoyo on 1 1. xi. 29. Another pair of eggs were found in a 
dried leaf, that is to say one egg was securely held by the leaf while the 
other, adhering to it, projected into space at a height of five feet from 
the ground. On the same day newly hatched young Avere seen on 
bananas. 

Habitat. Though the Bagamoyo series were collected on bananas, 
others were seen on palms to the west of the town; the ]\Iorogoro 
gecko was also taken off a banana plant, but at Bangwe, north of 
Ujiji, these geckos were seen running over rocks almost at the water's 
edge. The species has twice been recorded from Lake Tanganyika. 

Lygodactylus picturatus picturatus (Peters) 

Plate l,fig. 1 

Hemidactylus picturatus Peters. 1870, Monatsb. Akad. Wiss. Berlin, p. 115: 

Zanzibar. 
Lygodactylus picturatus Boulenger, 1885, Cat. Lizards Brit. Mus., 1, p. 161: 

Magiba, Pangani, Tanganyika Territory. 
Lygodactylus manni Loveridge, 1928, Proc. U. S. Nat. Mus. 72, Art. 24, pp. 1-2, 

pi. i: Saranda, Ugogo, Tanganyika Territory. 
Lygodactylus picturatus picturatus Loveridge, 1929, L^. S. Nat. Mus. Bull. 151, 

p. 46: Localities in Kenya Colony; Victoria Falls. 

3 & eggs (M. C. Z. 30498-501) Mainland opp. Mombasa. 29. x. 29. 
2 (M. C. Z. 30502-3) Changamwe, Kenya Colony. 31. x. 29. 
6 (M. C. Z. 30504-9) Saranda, Ugogo. 28. xi. 29. 
1 (M. C. Z. 30510) Bagamoyo. 11. xi. 29. 
50 (M. C. Z. 30511-23) Dar es Salaam. 5. xi. 29. 

Native name. Garomwe was the name applied to the male by some 
native children at Changamwe; when shown a female they said that 
it was kibibi, literallv "little wife" in Kiswahili. 

Affinities. L. manni was based on a single specimen from Saranda 
which had a peculiarly marked throat and differed somewhat in pro- 
portions of the head. Special search was made at Saranda on the 
present occasion and six topotypes secured, all in the Aicinity of the 
Indian shops by the station; no trace of the gecko could be found in 



loveridge: African herpetology 289 

the surrounding bush where L. (irofci occurs. It seems highly probable 
that L. manni is an artificial importation. 

The rich black and yellow coloring of the underparts quite surpassed 
that of coastal males but of the series only two had the gular markings 
of typical manni. ^Measurements of a long series of males from Dar es 
Salaam also reveal that Saranda specimens are within the range of 
variation. Under the circumstances I consider manni a strict synonym 
of picturafvs typica. 

In an attempt to define ph\sical characters of two well-marked 
color varieties I utilized a watchmaker's instrument for taking very 
fine measurements of the length and breadth of these gecko's heads. 
The results are gi\'en under each ^•ariety. 

Variation. Breadth of head is included in length from 1.3-1.7 times 
(only two with 1.7) and an average of 1.4, based on 25 geckos from 5 
localities; preanal pores 6-9, average 8.0 based on 11 males. 

Measurements. Largest male (No. 30518) measures 81 (41+40) 
mm. ; largest female 72 (38+34) mm. 

Breeding. Eggs taken on October 29th measured 5.5 x 6.5 mm. 
There were six pairs of these eggs in a hole in a palm trunk on which the 
adult geckos were taken, of these eggs two had hatched, one hatched 
on the 30th, the remainder were preserved. At Bagamoyo two eggs 
were found in a sunbird's empty nest. 

Habitat. At Changamwe on a coconut palm and a paupau stem, 
young ones were seen on the fence and paling surrounding the railway 
station; on a tree trunk on Momboni Road, Tanga; the series from Dar 
es Salaam were all obtained in one day by Salimu from Acacia trees in 
Main Avenue. Three were taken on a kengi tree at Saranda, the others 
on acacia. 

Lygodactylus picturatus var. on Mombasa Id. 

Plate 1, fig. 2 

Lygodactylus picturatus (part) Loveridge, 1920, Proc. Zool. Soc. London, p. 139: 
Frere Town; Mombasa; and Jilore, Kenya Colony; Loveridge, 1923, Proc. 
Zool. Soc. London, p. 941 : Frere Town and Kilindini, Kenya Colony. 

10 (M. C. Z. 30586-93) Kilindini, Mombasa Id. 28. x. 29. 

Distribution. A female w^as seen at Tanga. So far as is at present 
known, therefore, this color variety occurs along the coast from Jilore 
to Tanga; at several places in the same locality as typical yellow- 
headed picturatus though never on the same trees so far as my ex- 
perience goes. 



290 bulletin: museum of comparative zoology 

Affinities. A form intermediate between L. p. induratus and L. p. 
gutturalis, lacking the yellow head of the former and the gular chevrons 
of the latter though these are occasionally faintly indicated in females. 

Variation. Breadth of head is included in length from 1.3-1.5 times 
(only one with 1.5) and an average of 1.37 times, based on 18 geckos 
from 2 localities; preanal pores 8-10, average 9.1, based on 12 males. 

Measurements. Largest male (No. 30587) measures 81 (39+42) 
mm., largest female (No. 18537 Frere Town) measures 67 (37+30) 
mm. 

Habitat. These geckos are now quite common on the trees forming 
the avenue to Kilindini station. A few hours before leaving East 
Africa I went ashore and unaided captured a dozen alive and unin- 
jured. Unfortunately someone on board, probably a child, left the 
top of the vivarium open, so that eight escaped; the remainder did not 
survive the voyage. 

Lygodactylus picturatus var. on Ukerewe Id. 
58 & eggs (M. C. Z. 30536-60) Ukerewe Id., Lake Victoria. 10-12. vi. 30. 

Native name. Kihangalla (Kikerewe). 

Affinities. Another form intermediate between L. p. picturatus and 
L. p. gutturalis but nearer to the latter than is the Mombasa form for 
there are definite tendencies for the black of the throat in males to 
form chevrons. From the Mombasa form it differs not only in the 
head markings but in its larger size. 

Variation. Breadth of head is included in the length from 1.1-1.2 
times and an average of 1.7 times, based on 25 geckos from the above 
series; preanal pores 7-9 (except for an aberration in one of 4), average 
7, based on 15 males. 

Coloration. Noted in life, d^ . Above, grey mottled with black and 
having a row of light colored ocelli (sometimes black-edged) along 
either side of the vertebral line; on the side of head and neck are about 
three rows of interrupted, but very conspicuous, black stripes. Be- 
low, upper and lower labials china-white marked with jet-black prin- 
cipally along the buccal borders; throat deep velvety -black (in some 
specimens, particularly in young males, there is a tendency for the 
black to be arranged in A- formations) ; lower side of neck a rich orange, 
though Chinese-white in some fully adult males, extending almost to 
the fore arms, on either side of the neck some black speckling; a broad 
band along the length of the breast and belly, as also the under side 
of the limbs, is pale yellow; flanks on either side of this band, as also 



loveridge: African herpetology 291 

on the underside of the tail, greyish white deepening to grey towards 
the tip of the tail. 

9 . Above, rather more browTiish-grey than the male though 
mottled and ocellated as in that sex. Below, greyish-white except on 
the breast and the pelvic area including the under side of the hind 
limbs which are a very pale yellow. 

Young. In these the under side of the tail is frequently salmon-pink. 

Measuremetifs. Largest male (Xo. 30542) measures 90 (42+48) mm., 
largest female (No. 30544) measures 74 (37+37) mm. 

Breeding. Four clutches of eggs, of which one pair is separated, were 
obtained on June 12th and measured 6x7 mm. 

Enemies. Two were recovered from the stomach of a Spotted Wood 
Snake (Philothamnus s. semivariegatus) . 

Habitat. Some were taken on mango trees but they were more 
plentiful on the trunks of the imported Javan silk-cotton trees while 
a few were actually taken on buildings, a most unexpected place for 
members of this arboriphile genus. 

Lygodactylus picturatus gutturalis (Bocage) 

Hemidactylus gutturalis Bocage, 1873, Jorn. Sci. Lisboa, p. 211: Bissao, Portu- 
guese Guinea. 

Lygodactylus picturatus gutturalis Schmidt, 1919, Bull. Amer. Mus. Nat. Hist., 
39, p. 462: Garamba, Belgian Congo. 

54 (M. C. Z. 30561-85) Ujiji, Lake Tanganyika. 28. v. 30. 

Distributioti. This is the first record of the occurrence of this race 
in Tanganyika Territory though to be expected for it had been re- 
ported from the Ki\u Region by Sternfeld. 

Affinities. A well-marked color variant of L. picturatus character- 
ized by clearly defined chevron-shaped markings on the throat. 

Variation. Breadth of head included in the length from 1.1-1.2 
times (only 3 with 1.2) and an average of 1.1 times, based on 25 
geckos; preanal pores 7-8, average 7.5, based on 16 males. 

The whole series have been examined for broken tails and the 
majorit}' found to have them intact, all such, except on the very tip, 
have a row of single transversely enlarged subcaudals though here 
and there one of these scales may be divided, such abnormalities, how- 
ever, are unusual and do not form more than five per cent of the total 
scales on anyone tail; on the other hand in regenerate tails the majority 
of transversely enlarged scales are only about half the width of similar 
scales on the uninjured basal portion of the same tail and they exhibit 



292 BL'LLpTIN: MUSEUM OF COMPARATIVE ZOOLOGY 

a fairly high proportion of divided or small scales interspersed among 
the transversely enlarged ones. 

Coloration. Every gecko in the series has three dark che\Ton-shaped 
gular markings, sometimes there is a spot between the arms of the 
posterior, i.e. smallest, chevron, or in a very few the whole area be- 
tween the arms may be filled in so as to form a black triangular patch. 

Measurements. Largest male (No. 30561) measures 84 (42+42) 
mm., largest female (No. 30577) measures 72 (37+35) mm. 

Breeding. Some of the females hold large eggs but none was ready 
for laying. 

Lygodactylus angularis Giinther 

Lygodactyliis angularis Giinther, 1893 (1892), Proc. Zool. Soc. London, p. 555, 
pi. xxxiii, figs. 1-3: Shire highlands, Nyasaland. 

1 (M. C. Z. 30468) Tandala, Ukinga Mtns. 11. ii. 30. 
22 (M. C. Z. 30469-85) Madehani, Ukinga Mtns. 14-19. ii. 30. 
1 (M. C. Z. 30486) Nkuka Forest, Rungwe Mtn. 28. iii. 30. 

Distribution. This species, hitherto only known from Nyasaland, 
must now be added to the fauna of Tanganyika Territory. 

Xativc names. Linyarupancja (Kihehej; kamhiri (Kikinga); komaki- 
piki (Kinyakusa; the Banyakusa consider it the young of Agama 
atricollis). 

Variation. Upper labials 5-8, average of forty-eight counts almost 
7; lower labials 5-7, average 6; preanal pores of males 5-8, average 6. 
The transversely enlarged sul^caudals distinguish this fine gecko from 
L. fischeri, the only other rain-forest member of the genus which ap- 
proaches it in size. 

Coloration. The gular pattern of this species at once distinguishes 
it from all other East African members of the genus, roughly it might 
be said that the chevrons of L. p. gutturalis are inverted with their 
apices pointing towards the body, they are well shown in Giinther's 
plate. 

The Tandala male, in life, had the six gular lines on a lemon-yellow 
background, with the exception of the forearms the rest of the under- 
surface was rose-pink, brighter in the anal region. No other East 
African Lygodactylus has a pink ^•entral surface. 

Females from Madehani differed entirely from the males in being 
wholly lemon-yellow below without any rose-pink. The gular lines 
show considerable variation in detail though remaining characteristic. 

Measurements. The largest of nine males measures 84 (42+42) mm. ; 
the largest of fifteen females 81 (46+35) mm. 



loveridge: African herpetology 293 

Breeding. Eggs in various stages of development were present in 
all the females taken at Madehani, 14-19. ii. 30, the largest ova 
measured 8 mm. in diameter and were presumably about ready for 
laying; others held eggs of 7 and 6 mm. diameter on the 14th. Two 
eggs only are developed at a time. 

Diet. Stomachs of ten geckos were examined with the following 
results: — (1) Many beetles, including a Curculionid and Lampyrid, 
(2) beetles, (3) beetle, ant, (4) beetle larvae, spider, (5) beetle, two 
caterpillars, (6) beetle, two caterpillars of which the larger measured 
35 mm.! (7) hairy caterpillar, (8) big black ant, (9) ants, bug, (10) 
bug, spider. In addition two lizards held remnants of eggshell 
which I have little hesitation in saying w^re from Lygodactyli eggs, 
they were in the stomachs, not oviducts, and I feel certain that they 
were not snail shells. 

Parasites. Acarine parasites are present near the anus. 

Habitat. The first specimen was on the trunk of an eucalyptus tree 
immediately behind the kitchen building of the ruined German mission 
house at the end of the long avenue. 

Being a species unknoA\Ti to me I was thereafter continually on the 
search for more but found none until, shortly after our arrival at 
Madehani, I shot two on tree trunks flanking the road where it passes 
through big forest. These specimens were show^n to all local natives 
and one particularly bright lad responded by bringing in sixteen; six 
of these had discarded their tails but he had the good sense to bring 
in the tails as well. It should not be inferred from this that the species 
is particularly abundant at Madehani for though I was constantly on 
the lookout for this gecko I rarely saw more than one per day during 
the three weeks of our stay. One I shot on an iron telegraph pole; 
another, which I caught on a tree-trunk, was engaged in casting its 
skin. 

• Pachydactylus boulengeri Tornier 

Pachydactylus boulengeri Tornier, 1897, Kriechthiere Deutsch-Ost-Afrikas, 
p. 26, pi. ii, figs. 1-2: Tabora and Kakoma, Tanganyika Territory; Lov- 
eridge, 1923, Proc. Zool. Soc. London, p. 941 : Sagayo, Tanganyika Terri- 
tory. 

Elasmodadylus triedrus Boulenger, 1913, Revue Zool. Afr., 3, p. 104, text fig: 
Kikondja, Katanga, Belgian Congo; Loveridge, 1920, Proc. Zool. Soc. 
London, p. 140: Morogoro and Kongwa, Tanganyika Territory; 1923, 
Proc. Zool. Soc. London, p. 942: Suna, Tanganyika Territory; 1928, Proc. 
U. S. Nat. Mus., 73, Art. 17, p. 63: Saranda, Tanganyika Territory. 



294 bulletin: museum of comparative zoology 

1 (M. C. Z. 30594) Tanga. 2. xi. 29. 

1 (M. C. Z. 30595) Handa, Usandawi. 10. xii. 29. 

1 (M. C. Z. 30596) Mangasini, Usandawi. 11. xii. 29. 
5 (M. C. Z. 30597-601) Saranda, Ugogo. 28. xi. 29. 

5 (M. C. Z. 30602-5) Nyamkolo, Lake Tanganyika. 9. v. 30. 

2 (M. C. Z. 30606-7) Kasanga, Lake Tanganyika. 16. v. 30. 

3 (M. C. Z. 30608-10) Shinyanga, Usukuma. 3. vi. 30. 

In addition the following material from Tanganyika Territory in 
the Museum of Comparative Zoology has been employed for collecting 
the undermentioned data. 

1 (M. C. Z. 18265) Sagayo, Usukuma. 13. xi. 22. 
1 (M. C. Z. 18270) Kongwa, Ugogo. 21. iv. 17. 
. 2 (M. C. Z. 18271-2) Suna, Unyaturu. 8. x. 21. 
1 (M. C. Z. 22977) Mbala, Usagara. 26. ii. 23. 
1 (M. C. Z. 23045) Saranda, Ugogo. 19. vi. 26. 

Relations. The specimens from IVIorogoro and Kongwa referred to 
in the above citation, were determined for me as E. triedrus by Boulen- 
ger himself. The holotype being at the Congo Museum, Tervueren, 
we depended on the description and text figures. 

During the present expedition I obtained material from all round 
the type locality of P. boulengeri and southwards to Nyamkolo where 
there is an admixture of Katanga fauna (e.g. Chilorhinophis gerardi, 
Mabuya perrotetii, etc.) and one might reasonably expect to find 
E. triedrus. Nyamkolo geckos, however, are indistinguishable from 
those of Central Tanganyika. 

P. hoidengeri lacks, or at least I fail to find, a minute, concealed, 
retractile claw on the digits which is one of the alleged generic charac- 
ters of Elasmodactylus . Though triedrus was referred to that genus no 
definite claw appears in the enlarged drawing of the foot, nor is specific 
mention of a claw made in the very detailed description. As the de- 
scriptions and figures of the two species appear to be in entire agree- 
ment I have little doubt but that they represent a single species. 

Subsequently to \\Titing the above the Kongwa gecko was sent to 
Dr. Gaston de ^^'itte at the Congo Museum. He replies: "I have very 
carefully compared your specimen with the type of Elasmodactylus 
triedrus Boulenger. You are perfectly right. It is impossible to dis- 
cover retractile claws, but there are small sheaths which perhaps made 
it appear as if retractile claws existed. This character seems to have 
little importance and one should perhaps unite Elasmodactylus and 
Pachydactylus. Your identification of it with Pacliydactylus boulengeri 
is correct." 



loveridge: African herpetology 295 

Variation. Upper labials 8-11; lower labials 6-9; nostril pierced 
between the rostral, 1st labial and 3 or 4 small scales, in 12% of the 
series, however, the 1st labial is excluded from bordering the nostril; 
males with 7-10 preanal pores, average of ten males is 7.7. The species 
is very distinct from P. bibronii of which there are specimens from 
Igulwe and Dodoma, Ugogo in the Museum of Comparative Zoology. 

Measurements. Largest of ten males (No. 23045) measures 167 
(79+88) mm.; largest of fourteen females (No. 30608) measures 133 
(63+70) mm. The Tanga gecko is very young being only 47 (25+22) 
mm. 

Habitat. The Tanga specimen was taken among rejected mcuti 
thatching beside a hut; the Handa gecko beneath the bark of a dead, 
but still standing, tree beside the trail from Masiliwa to Handa; it al- 
ready lacked a tail when first seen. Apart from the Mangasini lizard 
which was brought in by a native, all the rest were found on the 
whitewashed walls of huts or houses. At night they emerge from the 
thatch or roof where they spend the day in complete concealment. So 
shy of a flashlight were these geckos at Saranda and Nyamkolo that 
they made for the eaves as soon as the light fell upon them, this 
timidity made it necessary to shoot them with a fine charge of dust 
shot which did not damage them in the least. 

Diet. Beetles were recovered from the stomach of a Saranda gecko. 



Phelsuma laticauda (Boettger) 

Pachydadylus laticauda Boettger, 1880, Zool. Anz., 3, p. 280: Nossi-Be, Mada- 
gascar. 

Phelsuma laticauda Loveridge, 1920, Proc. Zool. Soc. London, p. 139: Dar es 
Salaam; 1925, Proc. Zool. Soc. London, p. 74: Zanzibar; Barbour & Lover- 
idge, 1928, Mem. Mxis. Comp. Zool. 50, p. 146: Dar es Salaam and Baga- 
moyo, Tanganyika Territory. 

Breeding. Broken egg-shells of the Palm Gecko were seen at a 
height of six feet from the ground in crevices of the trunks of a coconut 
palm. As these are the first I have ever seen I assume that the species 
usually lays in the crowns of the palms where it lives. Though no 
specimens of this gecko were secured during our brief stay at Bagamoyo 
one of my collectors obtained a good series on a previous visit. 

Enemies. Remains of a Palm Gecko were found in the stomach of 
a Tiger Snake {Tarbophis semiannnlatns) . 



296 bulletin: museum of comparative zoology 

AGAMIDAE 

Agama hispida armata Peters 

Agama armata Peters, 1854, Monatsb. Akad. Wiss. Berlin, p. 616: Rios de 

Sena, Tete, Mozambique. 
Agama hispida var. distanii Loveridge {nee. Boulenger), 1923, Proc. Zool. Soc. 

London, p. 942: Tanganyika Territory records. 

3 (M. C. Z. 30611-2) Gulwe, Ugogo. 21. xi. 29. 
12 (M. C. Z. 30613-6) Unyanganyi, Turu. 3. xii. 29. 
2 (M. C. Z. 30617-8) Mangasini, Usandawi. 14. xii. 29. 
1 (M. C. Z. 30619) Ikikuyu, Ugogo. 21. xii. 29. 
1 (M. C. Z. 30746) Near Ikombo, N. Rhodesia. 6. v. 30. 

Correction. On my way out to East Africa in 1914 I collected some 
agama lizards at Delagoa Bay, Mozambique which were submitted to 
Dr. G. A. Boulenger and referred by him to distanii, a species which 
he had described from Pretoria and Rustenberg, Transvaal. In 1923, 
and in several subsequent papers, I referred all Tanganyika examples 
to hispida distanii using the Delagoa Bay specimens for reference. 
Recent examination of topotypes of distanti, however, has re\'ealed 
that the Delagoa Bay agamas (and consequently all my Tanganyika 
material) were not of that race but should be referred to Agama hispida 
arinaia. 

Variation. The ventrals are faintly keeled except in one male 
(No. 30613) where they appear to be smooth; the middle (3rd) toe 
is longer than the 4th except in two specimens (No. 30614) where the 
4th is slightly longer. Whether the ear-opening is larger than the eye- 
opening is an impossible character to decide upon where both are so 
much of a size, generally they are substantially equal, but if the verti- 
cal — as opposed to the horizontal — diameter of the ear-opening be 
taken then it is larger than the eye-opening. Preanal pores in the 
males 10-13. 

Measurements. Largest male 228 (90+138) mm.; largest female 196 
(81 + 115) mm. Both are from Gulwe and both are exceeded by 5 mm. 
in body length by Unyanganyi agamas with injured tails. Smallest 
agama (No. 30746) measures 75 (37+38) mm. 

Breeding. All the Unyanganyi females are heavy with eggs ready 
for deposition; in three lizards examined the eggs numbered 10, 12 and 
13 respectively, while they measured 15 x 10 mm., 13 x 9 mm., and 
12x8 mm. 

Parasites. Immature nematodes (Physaloptera sp.) were preserved 
from an Unyanganyi agama. 



loveridge: African herpetology 297 

Habitat. The Guhve specimens were taken on stumps in the desert 
two hundred yards south of the railway station; those from Unyan- 
ganyi were captured on baobab as well as smaller trees. 

Agama agama lionotus Boulenger 

Agama lionotus Boulenger, 1896, Proc. Zool. Soc. London, p. 214, pi. viii: 

southeast of Lake Rudolph, Kenya Colony. 
Agama agama lionotus Loveridge, 1929, U. S. Nat. Mus. Bull. 151, pp. 48 and 

53: Kenya Colony and Tanganyika Territory localities. 

4 (M. C. Z. 30620-3) Kilindini, Mombasa Id., K. C. 28. x. 29. 

Variation. jVIidbody scales rows in males 70, in females 69-73; 
preanal pores in males 11-14. 

Coloration. ]\Iale. Above, crown and sides of head gamboge or 
mustard yellow which extends as a gradually narrowing streak along 
the vertebral line to midbody, the rest of the vertebral streak to the 
tail is silvery; sides of body very dark navy blue upon which two 
parallel rows of large black blotches may be discerned; fore limbs 
metallic light blue; hind limbs metallic greenish-blue ringed with 
lighter as every third scale-row is almost white. Below, throat, in- 
cluding the gular skin, brick-red; shoulders metallic ultramarine; 
on the sides of the neck a small patch of black separates the yellow 
from the blue; body greenish-blue except for a narrow line from about 
midbody, this widens till it almost excludes the blue from the preanal 
region and from the undersides of the hind limbs and the lower surface 
of the tail. 

Measurements. Larger male measures 262 (112-[-150) mm.; larger 
female 231 (85+146) mm. 

Breeding. Both females were bloated with eggs which took up all 
available space in the body cavity extending forward to the chest. 
These eggs, which numbered 6 and 7 respectively, varied greatly in 
size according to the position into which they had been squeezed, 
three measured as follows: — 21 x 10 mm., 20 x 11 mm. and 19 x 12 
mm. 

Diet. Ants appeared to form the principal stomach contents but 
there were remains of beetles and a single millipede. 

Parasites. Nematodes were present. 

Habitat. On the outward voyage these lizards were abundant on 
Kilindini wharf where they basked upon the piles of girders and sought 
refuge from pursuit beneath them ; they also occurred along the fence 
of corrugated iron sheets which encloses the wharf and which was 



298 bulletin: museum of comparative zoology 

flanked with a rank growth of weeds and piles of rock. On the return 
voyage it was observed that the girders had been removed and the 
lizards were apparently not so plentiful. It is doubtful, however, 
whether the clearing up of the area will greatly affect them as they 
have taken to living on the walls of the warehouses opposite the station 
and retire beneath the roofs if molested. 

Agama agama mwanzae Loveridge 
Plate 2, fig. 4 

Agama lionotus var. mwanzae Loveridge, 1923, Proc. Zool. Soc. London, p. 945: 
Shanwa, Mwanza, Tanganyika Territory. 

19 (M. C. Z. 30624-35) Shinyanga, LTsukuma. 3. vi. 30. 
37 (M. C. Z. 30636-60) Mwanza, Usukuma. 6. vi. 30. 
26 and eggs (M. C. Z. 30661-85) Ukerewe Id., Lake Victoria, vi. 30. 

Distribution. Hitherto this race has not been recorded from outside 
the Usukuma country where it occurs at Mwadira and Sagayo in 
addition to the four localities given above. 

Native names. KiM (Kisukuma); bidungu (Kikerewe). 

Variation. Midbody scale-rows in fifteen males 72-84, in fifteen 
females also 72-84; preanal pores in fifty males 8-12, average 10, two 
males have supernumary rows of 9 and 2 pores respectively which were 
omitted from the count. 

Coloraiion. An exceptionally handsome male had two dark blue 
patches on the posterior portion of its rich magenta throat, this con- 
stitutes a quite abnormal variation. 

Measuremenis. The largest male (Shinyanga) measures 364 (129+ 
235) mm.; largest female (Mwanza) is 275 (115+160) mm. 

Brccdirig. At noon on June 11th, as I was walking along a sandy 
path on Ukerewe Island, I observed the head of a female agama pro- 
truding from a horizontally oval hole in the path; even as my eye fell 
upon her she darted out and as she did so I observed an egg roll back 
into the hole. A closer examination of the hole showed that its en- 
trance was about three inches wide by an inch in height, for the first 
three or four inches the burrow^ sloped steeply downwards so that its 
terminus was two and a half inches below the surface. Besides the 
egg which I had seen roll back, and which was lying on some loose sand, 
four other eggs were found packed in loose sand beneath an undercut 
ledge of soil which had not been disturbed. The eggs were white and 
measured 21 x 9 mm., the texture of their envelopes being of the same 
tough, parchment-like structure common to snakes' eggs. 

Parasites. These specimens showed the same heavy infestation of 



loveridge: African herpetology 299 

nematodes {Sfrongi/luris hrevicauduta and <S. ornata). that I have noted 
elsewhere. 

Enemies. Several eggs of this agama were found in the stomach of 
a Nilotic Monitor (J'aranus niloiicus) which I shot as it was basking 
on a promontory on which A. a. mwanzae were more abundant than 
anywhere else on those parts of the island w^hich I visited. 

Habitat. At Shinyanga these lizards were very plentiful on the 
group of rocks southwest of the railway station. On Ukerewe Island 
it seemed as if the agamas were less plentiful inland where the rocks 
did not provide them with so many fissures as along the shore. 

Agama agama turuensis Loveridge 
Plate 2, fig. 1 

Agama agama turuensis Loveridge, 1932, Bull. Mus. Comp. Zoo!., 72, p. 376: 
Unyanganyi, east of Singida, Tanganyika Territory. 

36 (M. C. Z. 30686-30710) Unyanganyi, Turu. 3^. xii. 29. 
24 (M. C. Z. 30711-30735) Mangasini, Usandawi. 12. xii. 29. 

Distrihution. Both geographically and in its gular markings this 
form is intermediate between the northern A. a. lionotus and more 
southerly A. a. dodomae. The former has a plain red throat and the 
latter a red centre surrounded by a broad blue band of which the 
marking in turuensis is obviously the beginning for three of the sixty 
specimens listed above show faint traces of this marking which is 
strongly developed in dodomae though in blue instead of black. 

Affinities. Since describing this form I have had the opportunity 
of examining one of the male cotypes of Agama elgonis Lonnberg 
which is undoubtedly so nearly related to turuensis that the latter may 
have to be ultimately placed in the synonymy of .4. a. elgonis. On 
the basis of the available data (7 topotypes of elgonis and 60 of turuen- 
sis) they may be distinguished as follows : 

Midbody scale-rows 80-90; preanal pores in males 14 A. a. elgonis 

Midbody scale-rows 72-82; preanal pores in males 

9-14, average for thirty-four males 11.3 A. a. turuensis 

It is important to note that the gular markings are identical though 
elgonis comes from an altitude of 8,000 feet and turuensis only circa 
4,500 feet. 

Diet. Ants and termites are present in the stomachs. 

Parasites. Nematode worms {Oochoristica theileri) in stomachs. 

Habitat. Usually found on rocks but a few were living in holes of 
the baobab trees. 



300 bulletin: museum of comparative zoology 

Agama agama dodomae Loveridge 

Plate 2, fig. 2 
Agama lionotus var. dodomae Loveridge, 1923, Proc. Zool. Soc. London, p. 944: 
Dodoma, Ugogo, Tanganyika Territory. 

10 (M. C. Z. 30736-40) Dodoma, Ugogo. 23. xii. 29. 

Distribution. Also seen at Kilimatinde and Saranda. The above 
series of ten males were shot between Dodoma and Kikuyu, the latter 
being less than two miles from Dodoma. The race seems to occur 
south of Dodoma nearly to the Ruaha River as many were seen, though 
none collected, on the motor run from Dodoma to Iringa. 

Variation. Preanal pores 10-12, a^•erage 11.4, which agrees with 
that of the type series of 35 specimens. 

Parasites. Nematodes {Stronc/i/luris gigas and Strongyluris ? ornata) 
were recovered from the stomachs of these agamas. 

Agama agama ufipae Loveridge 

Plate 2, fig. 3 
Agama agama ufipae Loveridge, 1932, Bull. Mus. Comp. Zool., 72, p. 377: near 
Kipili, Ufipa, Tanganyika Territorj'. 

5 (M. C. Z. 30741-5) Kipili, Ufipa. 19. v. 30. 

Coloration in life. — 9 paratype. Above, crown of head nearly black 
with sharply defined white spots; back pale buff , Jieavily vermiculated 
with black-edged, vermilion lines, black-edged white spots and sepia- 
colored blotches; fore limbs grey vermiculated with brown; hind limbs 
and tail sandy -buff vermiculated with sepia brown and black. Below 
pure white, except on the throat which is dusky in the centre and al- 
most cream on the sides. 

Parasites. Nematodes {Plu/salopfera sp.) were present in the 
stomachs. 

Agama atricollis Smith 

Agama atricollis A. Smith, 1849, IIlus. Zool. S. Africa, 3, Appendix, p. 14: 
Natal, South Africa. 

1 (M. C. Z. 30747) Matema, near Mwaya. 28. ii. 30. 
1 (M. C. Z. 30748) Mwaya, Lake Nyasa. 1. iii. 30. 
3 (M. C. Z. 30749-51) Tukuyu, Rungwe. 13. iii. 30. 

1 (M. C. Z. 30752) Ilolo, Rungwe. 28. iii. 30. 

9 (M. C. Z. 30753-4) Entebbe, Uganda. 27. vi. 30. 
3 (M. C. Z. 30755-6) Kampala, Uganda, vi. 30. 

2 (M. C. Z. 30757-8) Jinja. Uganda. 30. vi..30. 

13 (M. C. Z. 30759-60) Mabira Forest, Uganda. 1. vii. 30. 



loveridge: African herpetology 301 

Distribution: Also seen on trees in King Street, Tanga and near the 
Sise River on the Tukuyu-Abercorn Road, Northern Rhodesia. 

Native names. Kanakipiki and komakipiki (Kinyakusa, but these 
names are applied to the geckos Lygodactylus and Ilcmidactylus which 
the Banyakusa believe to be the young of the agamas). 

Variation. The nasal opening is actually below the sharp edge of 
the canthus; the ventrals in all these specimens show some keeling, 
in some scarcely distinguishable, in others very strong but uncorre- 
lated with geographical distribution; males possess two and even three 
rows of preanal pores, in cases where three occur the anterior is ill- 
defined, the posterior row has from 10-12 pores, the second 8-13. 

Measurements. The largest male (Mwaya) measures 345 (151 + 194) 
mm. the largest female (Entebbe) 285 (110+175) mm. 

Breeding. Two of the Mabira Forest agamas were examined and 
found to hold 22 and 21 eggs respectively, these measured 14 x 9 mm. 
and were apparently ready for laying. 

Diet. Beetles in the Ilolo specimen. One captive agama was seen 
to swallow a yellow migratory locust which came on board ship when 
in the Red Sea. 

Parasites. Nematodes {Strongyluris ornata) were preserved from 
Matema and Entebbe specimens. 

ZONURIDAE 

ZoNURUS Ukingensis Loveridge 

Plate 3, fig. 2 
Zonurus ukingensis Loveridge, 1932, Bull. Mus. Comp. Zool., 72, p. 378: 
Tandala, Ukinga Mtns., southwestern Tanganyika Territory. 
Type (M. C. Z. 30761) Tandala, Ukinga Mtns. 11. ii. 30. 

Affinities. In addition to the diagnostic characters already pub- 
lished it might be added that it is obviously not conspecific with 
tropidosternum, the only other species occurring in Tanganyika Terri- 
tory or any of those forms which ha\'e granular skin showing between 
the lateral scales {vide. Nieden, 1913, Mitt. Zool. Mus. Berlin, 7, p. 71, 
and Loveridge, 1920, Proc. Zool. Soc. London, p. 143). Moreover the 
shape of its frontonasal is entirely different from that of tropidosternum 
and that scale is separated from the rostral in the new form, not by a 
narrow suture of the nasals as sometimes occurs in tropidosternum , 
but by the whole breadth of the nasals. The complete, unpublished, 
description of the type follows. 

Description, cf. Head not depressed, longer than broad; head 
shields rugose; frontonasal as broad as long, separated from the rostral 



302 bulletin: museum of comparative zoology 

by the nasals which are broadly in contact; nostril pierced in the 
infero -lateral edge of a large nasal separated from the first labial by a 
narrow rim, prefrontals and a large preocular; prefrontals separated 
on the middle line by a small shield lying between the frontal and the 
frontonasal (apparently split off from the latter and possibly an ab- 
normality); frontal narrow, its lateral sides almost parallel, twice as 
long as broad; a pair of postfrontals separating the frontal from the 
anterior parietals; an occipital scale in the middle of a square formed 
by a pair of anterior and a pair of posterior parietals ; 4 supraoculars ; 
3 supraciliaries, the anterior very long; lower eyelid scaly and opaque; 
a large preocular; no loreal; 4 suboculars; 5 upper labials, the last 
tliree keeled; 5 lower labials, the last three keeled; a large mental 
followed by 5 pairs of chin-shields, the first pair in contact, the rest 
widely separated by small, but very strongly keeled, gular scales 
which are mucronate posteriorly. 

Dorsal scales large, their prominent keels forming raised ridges 
along the back, 31 scales between parietal and base of tail; flank scales 
rounded and strongly mucronate; 28 midbody scale-rows of which 
10 constitute the ventral series; ventrals obtusely keeled in 30 longi- 
tudinal rows exclusive of the row of preanals in which the middle pair 
are enlarged; 16 femoral pores. Tail with 13 whorls of scales, the dor- 
sal ones with enormously developed spines. 

Ch.^iaesaura miopropus Boulenger 

Chamaesaura miopropus Boulenger, 1894, Proc. Zool. Soc. London, p. 732: 
Fuambo, Nyasaland; Sternfeld, 1911, Mitt. Zool. Mus. Berlin, p. 385: 
Livingstone Mountains, Tanganyika Territory. 

1 (M. C. Z. 30762) Dabaga, Uzungwe. Mtns. 1. i. 30. 
1 (M. C. Z. 30763) Ipemi, Uzungwe Mtns. 8. i. 30. 
1 (M. C. Z. 30764) Ihenye, Ukinga Mtns. 8. ii. 30. 
1 (M. C. Z. 30765) Tandala, Ukinga Mtns. 11. ii. 30. 
1 (M. C. Z. 30766) Igale Pass, Poroto Mtns. 30. iv. 30. 

Distribution.. The Ukinga Mountains are the northern portion of 
the Livingstone Range; these new records link up the two earlier ones 
and extend the distribution to the northeast. 

Native navies. Xyoka lusagalla (Kihehe); nunduswa (Kikinga). 

Variation. To the original description the following may be added. 
Midbody scale-rows 24-26; upper labials 5-6; lower labials 4-5; 
femoral pores 1-2. 

Measurements. The largest lizard (Ihenye) measures 455 (106+349) 



loveridge: African herpetology 303 

mm. , and the smallest (Ipemi) only 157(42+111) mm. The length from 
snout to vent is included in the total length from 3.6 to 5.5 times and 
probably has some sexual significance. 

Breeding. The Igale female held a round egg measuring 7 mm. in 
diameter. 

Diet. An examination of the stomachs of the three largest lizards 
revealed (1) a black field cricket, (2) black field cricket and what ap- 
peared to be a beetle larva, (3) grasshopper and caterpillar. 

Habitat. Taken in long grass through which they travel with great 
speed and when on the move are indistinguishable from snakes. 

VARANIDAE 
Varanus niloticus (Linnaeus) 
Lacerta nilotica Linnaeus, 1766, Syst. Nat., ed. 12, p. 369: Egypt. 

1 (M. C. Z. 30769) Bagamoyo. 9. xi. 29. 

4 (M. C. Z. 30770) Mwaya, Lake Nyasa. 1-8. iii. 30. 

4 (M. C. Z. 30771) Kasanga, Lake Tanganyika. 16. v. 30. 

1 (M. C. Z. 30772) Mwanza, Lake Victoria. 6. vi. 30. 

1 (M. C. Z. 30773) Ukerewe Id., Lake Victoria. 13. vi. 30. 

Distribution. The foregoing were collected only for record. Nilotic 
Monitors were seen at Kilipi, L'jiji and Entebbe. 

Native name. Mhulu (Kinyakusa). 

Breeding. In the oviducts of the Ukerewe specimen were 21 eggs, 
each measuring 43 x 30 mm. 

Diet. The stomach contents were as follows: — (1) A string of what 
appeared to be snake's eggs, this opinion being strengthened by the 
presence of transverse ventral scutes, this monitor also held nineteen 
large slugs {Eleutherocaulis hrevis) Bagamoyo; (2) four snails at 
Mwaya; (3) grasshoppers, beetles and a crab's claw at Kasanga; (4) 
cockroaches, cricket and a large spider at Mwanza; (5) eggs of Agama 
agama mwanzae on Ukerewe Island. 

Parasites. The Bagamoyo monitor had spotted ticks {Aponomrna 
exornatum) about the anus, and brown ones of the same species in the 
armpits and elbow joints. Short worms {Tanqua tiara) in the stomach 
and enormously long ones {Stronglurus brevicaudata and S. ornata) in 
the mesentery. A 9 Filarioidea and tapeworms in a Mwaya specimen. 
Several lots of cestodes and nematodes ( T. tiara and Duthiersia fim- 
briata) as well as a 9 Oxyuroid were preserved from the Kasanga 
series. Tanqua tiara in the Ukerewe Monitor. 



304 bulletin: museum of comparative zoology 

Enemies. The tail of a Kasanga monitor was dead and dried, the 
suggestion was that it had been bitten by an aquatic cobra (Boulen- 
gerina a. stormsi). 

AMPHISBAENIDAE 

Amphisbaena mpwapwaensis Loveridge 

Plate 3, fig. 1 

Amphisbaena mpwapwaensis Loveridge, 1932, Bull. Mus. Comp. Zool., 72, 
p. 378: Mpwapwa, Ugogo, Tanganyika Territory. 

Types, cf 9 (M. C. Z. 30767-8) Mpwapwa, Ugogo. 22. xi. 29. 

Habitat. These specimens were taken by digging in dry earth be- 
neath a fallen tree close to a stream which meanders past the front of 
the new Veterinary Headquarters office built in 1929. 

LACERTIDAE 

Nucras boulengeri boulengeri Neumann 

Nucras tessellata Tornier (nee. A. Smith), 1897, Kriechthiere Deutsch-Ost- 

Afrikas, p. 39: South shore Victoria Nyanza. 
Nucras delalandii Tornier {nee. Milne-Edwards), 1900, Zool. Jahrb. Syst., 12, 

p. 593: Kakoma, Tanganyika Territory. 
Nucras boulengeri O. Neumann, 1900, Ann. Mag. Nat. Hist. (7), 5, p. 56: 

Lubwa's, Usoga, Uganda. 
Nucras eniini Boulenger, 1907, Ann. Mag. Nat. Hist. (7), 19, p. 488: Southern 

shore of Victoria Nyanza, Tanganyika Territory. 
Nucras ukerewensis Bolkay, 1909, Archivum Zool. Budapest, 1, p. 13, figs.: 

Shirati, Tanganyika Territory. 

55 (M. C. Z. 30774-85) Unyanganyi, Turu. 4. xii. 29. 
2 (M. C. Z. 31003-i) Masiliwa, f m-u. 9. xii, 29. 
1 (M. C. Z. 30786) Mangasini, Usandawi. 16. xii. 29. 
1 (M. C. Z. 30787) Shinyanga, Usukuma. 3. vi. 30. 

Distribution. East African records of this lizard are much involved; 
after an era in which attempts were made to identify them with South 
African members of the genus a decade followed in which specimens 
from the North, South and P^ast shores of Lake Victoria received 
names. I have visited all three type localities presuming that "Lub- 
was" means Lubwa's village near Port Lubwa, Usoga and there is 
nothing in the topography which would lead one to expect difl^erentia- 
tion of the fauna. 

The present range of A', b. buulcngcri, as understood by me, is from 
the East bank of the Nile at its source to Eldama Rner in Kenya 



loveridge: African herpetology 305 

Colony southeast to Mt. Kilimanjaro then due south to Kilosa where 
it meets with X. b. kilosac, thence due west to Kakoma and north 
through Tabora and Shinyanga to the South shore of Victoria Nyanza. 

Affinities. The coloring of N. h. boulengeri is so very distinct from 
that of iV. tcssellata that it was a source of surprise to me how records 
of the latter kept recurring in the literature even after the descriptions 
of boulengeri and cmini had appeared. A few years ago a series of 
specimens from the Loita Plains, Kenya Colony were received which 
differ from all Tanganyika examples that I have seen by possessing 
ocelli along the flanks and an olive ground color instead of a sandy 
or red-brown. I cannot help feeling that the records of tesseUata are 
based on lizards of this type. There do not appear to be any scale 
characters on which they can be separated, an apparent average num- 
ber of femoral pores may well disappear when larger series are avail- 
able. Specimens from Bissel, Kenya Colony and the adjacent Longido 
country are somewhat intermediate though lacking ocelli. Until more 
material is available it seems inadvisable to recognize this race, it 
may be that the name ukereivcnsis, which Boulenger considered a 
synonym of his enmii, might be used for this form. 

Boulenger, at the time of the writing of his Monograph of the 
Lacertidae, had very little East African material of the genus Nucras. 
N. boulengeri he only knew from the description and attempted to 
keep it distinct from emini on a supposed difference in the relative 
head lengths. The Unyanganyi series alone shows that the number of 
times which the head is included in the length from snout to vent is of 
no taxonomic importance in differentiating East African forms. The 
following figures are based on an examination of 50 boulengeri from a 
dozen different Tanganyika localities, 5 of the Loita Plains variety 
and half-a-dozen topotypic Kilosa lizards. 

N. boulengeri boulengeri 4.1-5.1 times 
iV. boulengeri var. (Loita) 4.1-4.9 
N. boulengeri kilosae 4.2-4.6 

At present I distinguish these forms l^y the following key: — 

Dorsal scales smooth 1. 

Dorsal scales keeled; body dark brown with richer 
and more abundant color pattern than the typi- 
cal form, includes a distinct light vertebral line . . N . b. kilosae 

1. Femoral pores average 11.5 (range 11-15); back 
light sandy -brown, uniform or finely speckled 
with black, a vertebral line usually absent, or if 
present, indistinct, except in the young N.b. boulengeri 



306 bulletin: museum of comparative zoology 

Femoral pores average 10 (range 9-12); back 
olive or olive-brown, much linear spotting of 
black and white (the latter formed from broken 
white lines) on back, with ocelli (white, ringed 
with black) on the flanks, more definite in some 
than in others N.b. var. 

Variaiion. The following data is based on 50 specimens from a 
dozen localities in Tanganyika, only 12 of the Unyanganyi series have 
been utilized. Dorso-lateral scale-rows across midbody 35-49, 40-49 
except for three lizards from the vicinity of Tabora and Suna; trans- 
verse rows of ventrals 28-34; longitudinal rows of ventrals 8; lamellar 
scales under 4th toe 17-24; femoral pores 9-15. 

Coloration. In addition to their more vivid dorsal coloring the 
young have bright red tails. 

Measurements. Largest intact specimen (Gwao's village) 190 (60+ 
130) mm., but four with reproduced tails surpass this in length from 
snout to vent being 67, 66, 66 and 61 mm. respectively. The smallest 
(Mukwese) lizard measures 98 (35+63) mm. 

Diet. The stomach contents of a dozen Unyanganyi lizards were : — 
(1) Cricket, (2) cricket, (3) cricket, (4) beetle, (5) grasshopper, 
termites, spiders, (6) termites, (7) termites, millipede annuli, stone, 
(8) termites, ants, (9) warrior termite, ant, (10) fly, spider, stone, (11) 
three spiders, (12) nil. 

At Masiliwa, following a heavy shower between 4 and 5 p.m. on 
December 9th, termites started flying in great numbers. After sunset, 
when already dusk, I picked up two of these lizards so gorged with 
termites that they scarcely tried to escape. One indeed ran slowly to 
its hole and inserted its head but could not get its body in; the hole 
being vertical, the sight was rather amusing, the lizard alternately 
WTiggling hard, then resting quiescent, after which I picked it up ! 

Enemies. On a path near Maji Malulu I disturbed a sandsnake 
{Psammophis biseriatus) swallowing one of these lizards whose head 
was already in its mouth. Two others were recovered from the 
stomachs of Psammojjhis subtaematu.s at Unyanganyi. 

Latastia johnstonii Boulenger 

LatastiajohnstoniiBoulenger, 1907, Ann. Mag. Nat. Hist. (7), 19, p. 392: Nyika 
Plateau, Nyasaland; Boulenger, 1921, Monogr. Lacert., 2, p. 16: Localities 
in Mozambique, Nyasaland, Southern Rhodesia and Tanganyika Territory. 

Latastia siebe7iroclii Nieden {nee. Tornier?), 1913, Mitt. Zool. Mus. Berlin, 7, 
p. 77: Tabora, Tanganyika Territory: Eldama River, Kenya Colony. 



loveridge: African herpetology 307 

4 (M. C. Z. 30788-91) Saranda, Ugogo. 28. xi. 29. 
12 (M. C. Z. 30793-800) Unyanganyi, Turu. 3. xii. 29. 
2 (M. C. Z. 30801-2) Mangasini, Usandawi. 16. xii. 29. 

Distribution in Tanganyika Territory. I have previously recorded 
this lizard from Morogoro; Tindiga near Kilosa; Manyoni; Tabora; 
Nyambita; Sagayo and Bukoba. 

In 1905 Tornier described as Eremias siebenrockii a lizard from Porto 
Novo, Dahomey, West Africa which in 1913 Nieden showed was a 
Latastia and with which he identified a lizard from Tabora and four 
from Eldama Ri^'er. The East African records are undoubtedly refer- 
able to johnstonii whose occurrence in German East Africa was un- 
known at the time Nieden wrote. L. johnstonii is abundant at Tabora 
and as we know that it occurs alongside Nucras b. boulengeri in many 
places in Tanganyika there is no reason to suppose that there is any 
question of the Eldama records, for N. b. boulengeri has already been 
recorded from that locality. 

Affinities. Boulenger separates the two species as follows: — 

No gular fold; edge of collar serrated; 13-16 femoral 

pores on each side johnstonii 

A gular fold; edge of collar even; 10-14 femoral pores 

on each side sicbenrocki 

actually the type had 11, the Eldama lizards 10-13 and the Tabora 
lizard 14. 

The character of the gular fold is difficult to decide, undoubtedly 
East African lizards have no fold though it is indicated; while the edge 
of the collar is serrated normally, in the Mangasini lizards it might be 
said to be even ; the Eldama specimens negative the alleged difference 
in the number of femoral pores. Nieden states very definitely that 
there is no difference in markings. 

Should siebenrockii prove to be synonymous with johnstonii it would 
have to take precedence. One wonders if the type of the former really 
did come from Porto Novo but until that is settled it would be dan- 
gerous to assume identity between two species in such widely separated 
localities. 

Variation. Basing the count on 25 lizards from ten Tanganyika 
localities as listed above. Midbody scale-rows 39-55; femoral pores 
12-17, with an average of 14, if the Eldama specimens are included, 
then 10-17. 

Measurements. The largest (No. 30802) measures 182 (62-M20) 
mm., though several have tails longer by 20 mm. 

Diet. Stomach contents of eight Unyanganyi lizards were composed. 



308 bulletin: museum of comparative zoology 

in every instance, of termites, some were of a very large species; in 
addition to the termites one lizard appeared to have eaten an antlion 
larva. On being caught a Mangasini lizard disgorged two antlion 
larvae. 

Habitat. This species favors the sandy thorn-bush steppe which, 
though at a high altitude, is often very hot during the day. These 
lizards were fairly jplentiful at Saranda and Unyanganyi where they 
were collected in desiccated mbiu/icr and sandy gardens. At Mangasini 
it was interesting to note its occurrence alongside its big relative, 
L. longicandata revoili, but while rcvoili was plentiful, johnsto7iii was 
scarce. It was surprising that no examples were collected between 
Central Tanganyika and the type locality — Nyasaland. 

Latastia longicaudata revoili (Vaillant) 

Eremias revoili Vaillant, 1882, Miss. Revoil Pays Comal., Rept., p. 20, pi. iii, 

fig. 2: Somaliland. 
Latastia longicaudata var. revoili Boulenger, 1921, Monog. Lacertid., 2, p. 30. 

1 (M. C. Z. 30792) Kilimatinde, Ugogo. 29. xi. 29. 
16 (M. C. Z. 30793-30810) Mangasini, Usandawi. 12-16. xi. 29. 

Distribution. Other Tanganyika material used in this study is from 
Mtali's village, Mkalama; Bahi and Dodoma, Ugogo. 

Variation. Midbody dorso-lateral scale-rows 54-70; transverse 
ventral rows 28-32; femoral pores 6-11 with an average of 8.7 pores 
for 52 counts. 

Measurements. The largest (No. 30807) measures 317 (100-|-217) 
mm. ; the smallest 102 (32+70) mm. was taken at Dodoma on May 14, 
1926. 

Diet. At Kilimatinde I watched one pick up a bleached fragment of 
Achatina shell twice, dropping it each time. 

Parasites. Nematodes {Strongyluris brevicaudata) were present in a 
Mangasini lizard. 



Ichnotropis bivittata Bocage 

Ichnotropis bivittata Bocage, 1866, Jorn. Sci. Lisboa, 1, p. 43: Duque de Brag- 
anga, Angola: Boulenger, 1921, Monogr. Lacertid., 2, p. 183: French Congo; 
Belgian Congo and Angola. 

2 (M. C. Z. 30836-7) Ipemi, Uzungwe Mtns. 7. i. 30. 
Distribution. The above constitute the first records for the occur- 



loveridge: African herpetology 309 

rence of this West African species in East Africa. Salimu, who assisted 
me in catching them, reported seeing one which escaped him at Tan- 
dala, Ukinga Mountains on 11. ii. 30. 

Variation. After careful comparison with a female from Caconda, 
Angola in the collection of the Museum of Comparative Zoology, I 
can see no grounds for separating the Tanganyika specimens from the 
Angolan. The type of /. tanganicana Boulenger was examined in 
London and is quite distinct. 

Midbody scale-rows (including ventrals) 34-37; transverse ventral 
rows 24; gular series 24; femoral pores 12-13. 

Coloration in life. Realizing that this beautiful lizard was new to 
the Tanganyika fauna, a detailed description of its coloration was 
made in the field. Above, head deep brown; back olive, on either side 
of a faintly indicated, light vertebral line are a series of bright chestnut- 
brown squarish blotches whose outer edges are touched with black and 
sometimes a little white, anteriorly these blotches tend to coalesce; 
a light (anteriorly it is tinged with yellowish) dorso-lateral liiie has its 
origin near the last supraocular and disappears on the base of the tail; 
below it is another series of blotches which are rather more black than 
chestnut-brown and having the appearance of ocelli by reason of a 
bluish-white central spot in each ; a white band along the upper labials 
becomes bright yellow behind the eye, passes across the ear-opening 
and (beneath the black blotches) along the flank to the hind limb, 
having become whiter between the fore and hind limbs ; it is bounded 
below by a vermilion line commencing on the lower labials and passing 
along the flank to the hind limb but interrupted by the fore limb. 
Below, china-white except the regenerated portion of the tail which is 
brown. 

Measurements. The larger of these two males is 143 (59+84) mm. 

Diet. A cricket was in the stomach of one of these lizards. 

Defence. When captured one gave a faint squeak or chirp, both 
gaped widely showing the scarlet edges and black interstices of their 
mouths. 

Habifat. Both were taken running about in short grass on either 
side of the path on the northern ascent to the village. 



IcHNOTROPis SQUAMULOSA Peters 

Ichnoiropis squamulosa Peters, 1854, Monatsb. Akad. Wiss. Berlin, p. 617: 
Tete, Mozambique; Peters 1883, Reise nach Mossamb., 3, p. 49, pi. viii, 
fig. 2; Boulenger, 1921, Monogr. Lacertid., 2, p. 191. 



310 bulletin: museum of comparative zoology 

2 (M. C. Z. 30839-40) Unyanganyi, Turu. 4. xii. 29. 
1 (M. C. Z. 30838) Kitungulu, Urungu. 15. v. 30. 

Distributio7i. These records extend the range of this most northerly 
representative of the genus much further north as the only two locali- 
ties in Tanganyika Territory from which it is know^n are Kakoma and 
the Makonde Plateau. 

Variation. Midbody scale-rows (including ventrals) 46-54; trans- 
verse ventral rows 26-28; gular series 28; femoral pores in adult 13, 
indistinctly developed in the two young from Unyanganyi but ap- 
parently 8-10, though the range is 13-16. 

Coloration. The upper aspect of the adult agrees well with Peters' 
fine colored plate; below it was lemon-yellow in life. The young 
have a light pinkish-white, lateral line which unites with its fellow 
posteriorly to merge into the straw-coloring of the tail, thus present- 
ing a very different appearance to that of the adult. 

Measurements. The adult female measured 191 (67+124) mm., a 
3-oung one 104 (37+67) mm. 

Breeding. The female recorded by Tornier from Kakoma was said 
to have large eggs in its ovary on May 6th, the female from Kitungulu, 
taken on the loth of the same month, holds 11 eggs approximately 
10 X 7 mm. 

Diet. In addition to a mealworm, the stomach of the adult was dis- 
tended with beetles. 

Habitat. At Unyanganyi found in dry thorn-bush steppe. Near 
Kitungulu, on the Kitungulu to Kasanga trail, three of these lizards 
were seen. So rapid were their movements that they were not ob- 
served until they had dashed off the path into the long grass and scrub 
which bordered the track. The one secured was shot in the early 
morning as it paused for a second before taking refuge in a thicket. 

Eremias spekii spekii Giinther 

Eremias spekii Giinther, 1872, Ann. Mag. Nat. Hist. (4), 9, p. 381 : Unyamwezi, 
Tanganyika Territory; Boulenger, 1921, Monogr. Lacertid. 2, 235; 

Eremias spekii spekii Loveridge, 1929, U. S. Nat. Mus. Bull. 151, p. 64: Locali- 
ties in Kenya Colony. 

1 (M. C. Z. 30811) Miritini, Kenya Colony. 30. x. 29. 
1 (M. C. Z. 30812) Dodoma, Ugogo. 25. xi. 29. 
46 (M. C. Z. 30813-31) Unyanganyi, Turu. 3-4. xii. 29. 
4 (M. C. Z. 30832-5) Mangasini, Usandawi. 13. xii. 29. 

Distribution. Also seen, though scarce, on Mombasa Island; the 



loveridge: African herpetology 311 

mainland opposite Kilinclini; and at Changamwe, all in Kenya Colony. 
Not common at ]\Iiritini, very abundant in the sandy stubble fields of 
Unyanganyi, Mangasini and at Kikuyu, near Dodoma. 

Native name. Lamhela (Kisandawi). 

J'ariation. Only two in the whole series had the subocular excluded 
from the lip (Xo. 30813 and duplicate) and then on the left side only. 

Parosifcs. Nematodes {Cxyuroidca sp. 9 ) were present in an Un- 
yanganyi lizard. 

Enemies. One was recovered from the stomach of a snake {Rham- 
phiophis rostratus) and another from a Psammophis suhtaeniatus. 

GERRHOSAURIDAE 

Gerrhosaurus major major Dumeril 

Gerrhosaurus major Dumeril, 1851, Cat. method, coll. Rept., Paris, p. 139: 
Zanzibar. 

Only one was seen during the expedition, this was among rocks on a 
hill in the centre of the town of Mwanza, Lake Victoria. 

Gerrhosaurus major zechi Tornier 

Gerrhosaurus maior zechi Tornier, 1901, Beiheft, Arch. Naturg., 67, p. 74: Kete 
Kratje, Togoland. 

30 (M. C. Z. 30841-55) Mangasini, Usandawi. 13. xii. 29. 
1 (M. C. Z. 30856) Dodoma, Ugogo. 23. xii. 29. 

Distribution. Recorded from the Belgian Congo by Schmidt (1919), 
from Morogoro and Dodoma in Tanganyika by Loveridge (1920) and 
from Kenya Colony (1929, U. S. Nat. Mus. Bull. 151, p. 66 for dis- 
cussion on status). 

Native name. Kinhotei (Kisandawi). 

Variation., Midbody dorsal scale-rows 17-20 longitudinal rows; 
transverse dorsal rows 32-36; longitudinal ventral rows 10; femoral 
pores 12-18; a single frontonasal separated from the rostral. These 
figures are based on 14 Mangasini and 2 Dodoma specimens. 

Coloration. This form can only be separated from major typiea by 
color but the two present a very different appearance in this respect. 
In life some zechi had blue throats and a lateral band of red. 

Yellow-brown above; white beneath G. m. major 

Buff above, the centre of every scale black; pinkish- 
white beneath, the centre of every scale (except gulars) 

dark brown presenting a striped appearance G.m. zechi 



312 bulletin: museum of comparative zoology 

Measuremenis. The largest measured 545 (225+320) mm. 

Diet. Every one of 22 ]Mangasini lizards examined were found to 
have fed upon winged termites. 

Parasites. All harbored nematodes (Physaloptera sp.), and some 
cestodes {Oochoristica zonuri). 

Habitat. These big girdled-lizards live among the piled-up rock 
masses of the kopjes which are like so many islands in the semi-desert 
thorn-bush steppe. As our arrival at Mangasini coincided with the 
breaking of the rains and the flighting of the termites the reptiles were 
probably more in evidence than at other seasons. To shoot them would 
have shattered their tails and our attempts to capture them merely 
made us feel foolish for the creatures crept into their ledges till they 
were out of reach and often lay there. I explained the position to the 
small Wasandawi boys who, with the aid of their dogs and padded 
arrows, promptly secured the above series in a little more than twenty- 
four hours. 



Gerrhosaurus flavigularis flavigularis Wiegmann 

Gerrhosaurus flavigularis Wiegmann, 1828, Isis, p. 379: "Africa merid. Krebs." 

2 (M. C. Z. 30857-8) Mangasini, Usandawi. 13. xii. 29. 

5 (M. C. Z. 30869-60) Ukerewe Id., Lake Victoria. 11. vi. 30. 

Distribution. Another was seen on the western bank of the Ru^■u 
River about twelve miles from Bagamoyo. 

Variation. Midbody dorsal scale-rows 22-24 longitudinal rows; 
transverse dorsal rows 57-62; longitudinal ventral rows 8; femoral 
pores 15-20; lateral scales keeled; prefrontals broadly in contact. 

Measurements. The largest measured 467 (170+297) mm. 

Parasites. A tick was found on the throat of a Mangasini lizard. 

SCINXIDAE 

Mabuya maculilabris (Gray) 

Euprepis maculilabris Gray, 1845, Cat. Liz. Brit. Mus., p. 114: West Africa. 

Mabuia maculilabris Boulenger, 1887, Cat. Liz. Brit. Mus., 3, p. 164, pi. ix, fig. 
2: West Africa; Ambriz, Angola; Angasija, Great Comoro Islands; Barbour 
& Loveridge, 1928, Mem. Mus. Comp. ZooL, 50, p. 157: West, Central and 
East Africa, 14 localities. 

1 (M. C. Z. 31002) Bagamoyo, 11. xi. 29. 
51 (M. C. Z. 30861-85) Mwaya, Lake Nyasa. 1-8. iii. 30. 



loveridge: African herpetology 



313 



1 (M. 


C. 


z. 


1 (M. 


c. 


z. 


1 (M. 


c. 


z. 


4 (M. 


c. 


z. 


1 (M. 


c. 


z. 


1 (M. 


c. 


z. 


1 (M. 


c. 


z. 


1 (M. 


c. 


z. 



30886) Nyamkolo, Lake Tanganyika. 9. v. 30. 

30887) Kitungulu, Urungu. 15. v. 30. 

30888) Kasanga, Lake Tanganyika. 16. v. 30. 
30889-92) Ujiji, Lake Tanganyika. 28. v. 30. 

30893) Ukerewe Id., Lake Victoria. 14. vi. 30. 

30894) Entebbe, Lake Victoria. 27. vi. 30. 

30895) Jinja, Lake Victoria. 30. vi. 30. 
30896-902) Mabira Forest, Uganda. 1. vii. 30. 



Native name. Ulusakatii (Kinyakusa for Mahuya). 

Variation. Midbody scale-rows 30-34; keels on dorsal scales 3-9, 
normally 5; supraciliaries 3-7, normally 5; prefrontals in contact in 17 
specimens, separated in 31, in one skink (No. 30901) this is due to the 
presence of an interprefrontal scale which also separates the frontal 
from the frontonasal, this is also the only skink with 3 supraciliaries ; 
supranasals in contact in 28 specimens, forming an "X" in two, sepa- 
rated by the rostral and frontonasal in 17. 

Arranging this data by locality, as was done in Barbour and Lover- 
idge's 1928 paper, it would read as follows: — 

Greatest 

Number head and Longest Mid-body Number Keels 

of body tail scale- of supra- on 

Locality skinks in mm. in mm. rows ciliaries scales 

Bagamoyo ... 1 78 135 31 5 9 

Mwaya 25 83 158 30-32 5-7 7 

Nyamkolo... 1 71 133 30 5 7 

Kitungulu ... 1 73 - 30 5 7 

Kasanga 1 64 106 32 6-7 7 

Ujiji 4 80 169 32-34 5 7-9 

ITierewe 1 83 1.32 32 5 7 

Entebbe 1 60 125 32 5 3-7 

Jinja 1 56 121 32 5-6 3-5 

Mabira 11 90 173 30-32 3-5 5-7 

This fairly long series was collected in the hope of throwing light 
on the relationship of M. maculilahris to M. comorensis; a study of the 
material leaves me in a greater quandary than ever; at most comorensis 
appears to be a race of maculilahris. Undoubtedly Amani comorensis 
average much larger than maculilahris, they are stouter and occasion- 
ally have a higher number of midbody scale-rows (34-38), the type 
from Comoro Islands had 36. Whether the size of the Amani skinks 
is due to abundant food and congenial climatic conditions seems pos- 
sible for the Central African maculilahris are almost as large. A really 



314 bulletin: museum of comparative zoology 

difficult problem awaits solution. Apparently Mahuya maculilabris is 
a skink that reacts readily to its environmental conditions and pro- 
duces color forms which are ill-defined when long series are available, 
yet are very striking and often of a characteristic type in a given 
locality. 

The Entebbe and Jinja specimens, though similar in color and 
markings to the skinks from the Mabira Forest, which lies about mid- 
way between Entebbe and Jinja, difi'er from all the others in the series 
in possessing 3 very strong keels, occasionally with an outer indistinct 
pair, to the dorsal scales. 

Having examined Boulenger's series of Ruwenzori M. maculilabris I 
find them specifically identical with the British Museum series of 
M. comorcnsis from Johanna Island. 

Coloration. The Bagamoyo skink is similar in color to Sternfeld's 
M. houlengeri from Makonde highlands and very different to all the 
others in the series except the Kitungulu specimen which approximates 
to it dorsally. 

The Mwaya series came from just across the lake to Old Langen- 
burg, type locality in part of M. vi. rohrbecki Sternfeld. 

The Ujiji specimens are undoubtedly M. m. major Sternfeld from 
the Central Lake Region and which is the best marked of all the forms 
geographically. In life the coloring was very rich, much reddish- 
orange above and lemon-yellow below. 

The throat and flanks of the skink from Ukerewe Island were tinged 
with magenta (not unlike the shade of Againa a. mivanzae which was 
plentiful on the island) while the breast and only the central line of the 
belly were yellowish. 

It might be added that the coloring of the INlAvaya, Nyamkolo, 
Kitungulu and Kasanga specimens was so similar to that of a Frere 
Town comorcnsis that I unhesitatingly referred them all to that species 
in the field, (vide. Proe. Zool. Soc. London, 1923, p. 956). 

Measurements. The largest specimen (Xo. 30896) measures 260 
(90+170) mm. 

Diet. In the stomach of one Mwaya skink was a snail, cockroach, 
two crickets and a caterpillar. 

Enemies. One recovered from the stomach of a Hissing Sand-Snake 
{Psammophis sibilans) at Mwaya. 

Habitat. The Bagamoyo lizard was collected on, or at the base of, a 
banana; another was seen outside its hole in the stem of a coconut 
palm. At Mwaya I took five among some dry thatching grass which 
had been piled on a trestle, showing these to the native children I 



loveridge: African herpetology 315 

offered the equivalent of one cent (U. S. currency, i.e. 5c Tanganyika) 
and had fifty skinks brought in the following day though I should have 
said that they were decidedly uncommon! At Kasanga they are not 
uncommon and by no means rare on the rocks at the water's edge 
where they occupy the niche usually filled by M. v. varia. They were 
not seen on the Bangwe rocks near Ujiji; the specimens from Ujiji 
were captured on the thatch of a native hut. The Jinja lizard lived in 
a hole in a tree trunk and I saw an M. striata basking at the same time 
upon the bole. When first seen the Nyamkolo skink was also basking 
on a hollow tree trunk but escaped me. Later I returned and was sur- 
prised to see a full-grown M. planifrons basking right beside the 
M. maculilabris, they were not separated by half-an-inch and the same 
shot killed them both. These instances of members of the same genus 
occupying the same territory is of no small interest. The series from 
Mabira Forest were taken on, or under, logs and in the grass at the 
edge of paths at the forest edge. 

Mabuya planifrons (Peters) 

Euprepes (Euprepis) planifrons Peters, 1878, Monatsb. Akad. Wiss. Berlin, 

p. 203, pi. ii, fig. 2: Taita, Kenya Colony. 
Mabuia diesneri Sternfeld, 1911, Sitzber. Ges. Naturf. Freunde Berlin p. 248: 

Kibwezi, Kenya Colony. 

5 (M. C. Z. 30903-7) Saranda, Ugogo. 30. xi. 29. 

2 (M. C. Z. 30908-9) Unyanganyi, Turu. 4. xii. 29. 

1 (M. C. Z. 30910) Mangasini, Usandawi. 13. xii. 29. 

4 (M. C. Z. 30911-3) Nyamkolo, Lake Tanganyika. 9. v. 30. 

Distribution. Xieden has recorded diesneri from Usumbura and 
Tabora and the present writer from Tabora, Izikisia, Ulugu and 
Ndogwe but under the earlier name of planifrons. 

Native name. Mbutlanga (Kisandawi). 

Variation. Midbody scale-rows 29-32; dorsal keels 3-5, usually 3 
with 5 anteriorly; ear lobules 3-4; frontal in contact with the 2nd and 
3rd supraocular; prefrontals in contact; postnasal in contact with the 
2nd labial in 4 specimens, and in contact on one side only in the re- 
maining 4; the toes of the adpressed hind limb reach the toes of the 
fore limb in the two largest skinks, in the others they reach to the 
WTist. 

Coloration. The two rows of irregular black spots present in the 
young are absent in the adults; the dark brown lateral band so well- 
defined anteriorly becomes indistinct about midbody and may be with 
or without white flecks. 



316 bulletin: museum of comparative zoology 

Measurements. The largest skink (No. 30908) measures 350 (125+ 
225) mm. 

Diet. Stomachs examined, held (1) an adult yellow migratory locust 
in an Unyanganyi skink, (2) flighting termites in the Mangasini 
reptile. 

Parasites. Nematodes (Stron.gyluris brevicaudata) were preserved 
from the stomach of an Unyanganyi specimen. 

Habitat. The Saranda skinks were all collected on tree trunks or 
fallen logs. The species is really rather common there though one 
might walk for days in the bush without seeing one. This is due to the 
sharp sight of the skinks which, seeing a person approaching from afar, 
dart round their tree trunk or take refuge in a hole. Further they are 
fond of basking head-downwards just about a foot from the ground, 
this is usually beneath the grass line and there is frequently a fissure 
or hole in the ground into which they can flee. One of the two skinks 
secured at Unyanganyi ran up a tree to escape. The Mangasini 
reptile was taken on a branch of an acacia shrub. 

Mabuya megalura (Peters) 

Euprepes (Mabuia) megalura Peters, 1878, Monatsb. Akad. Wiss. Berlin, p. 204, 
pi. ii, fig. 4: Taita, Kenya Colony. 

1 (M. C. Z. 20914) Bagamoyo. 11. xi. 29. 
11 (M. C. Z. 20915-6) Ilolo, Rungwe. 17-22. iii. 30. 
1 (M. C. Z. 20917) Ukerewe Id., Lake Victoria. 12. vi. 30. 
1 (M. C. Z. 20918) Nairobi, Kenya Colony. 5. vii. 30, 

Native name. Ulusakani (Kinyakusa for Mabuya). 

Variation. Midbody scale-rows 24-26, normal for the species. 

Measurements. The largest skink (No. 20917) measures 224 (60+ 
164) mm. 

Diet. The stomachs of seven Ilolo skinks were examined, one was 
empty, all the rest held spiders, the only other recognisable remains 
being of a very small grasshopper. It would appear as if this skink 
has a specialized diet as well as unusual habits. 

Habitat. The Bagamoyo specimen was caught in a swamp eight 
miles west of the town, another was seen climbing through the grass 
tops with astonishing agility. 

Mabuya varia varia (Peters) 

Euprepes (Euprepis) varius Peters, 1867, Monatsb. Akad. Wiss. Berlin, p. 20: 
Tete, Mozambique. 



loveridge: African herpetology 



317 



Mabuya varia Loveridge, 1929, U. S. Nat. Mus. Bull. 151, p. 74: Kenya aud 
Uganda localities — discussion. 

30911) Bagamoyo. 11. .\i. 29. 
30920-21) Unyanganyi, Turu. 3. xii. 29. 
30922) Handa, Usandawi. 10. xii. 29. 
30923-5) Mangasini, Usandawi. 12. xii. 29. 
30926-8) Dabaga, Uzungwe Mtns. 1. i. 30. 
30929) Ipemi, Uzungwe Mtns. 7. i. 30. 
30930-5) Kigogo, Uzungwe Mtns. 13. i. 30. 

30936) Lukungu, Ubena Mtns. 8. ii. 30. 

30937) Ihenye, Ukinga Mtns. 8. ii. 30. 

30938) Mangoto, Ukinga Mtns. 10. ii. 30. 
30939-42) Tandala, Ukinga Mtns. 11. ii. 30. 
30943-5) Madehani, Ukinga Mtns. 19. ii. 30. 
30946-9) IIolo, Rungwe, 22. iii. 30. 
30950-75) Igale, Poroto Mtns. 30. iv. 30. 
30976-7) Nyamkolo, Lake Tanganyika. 9. v. 30. 
30978-9) Kitungulu, Urungu. 15. v. 30. 
30980-1) Kasanga, Lake Tanganyika. 16. v. 30. 

30982) Kipili, Lake Tanganyika. 19. v. 30. 

30983) Ujiji, Lake Tanganyika. 28. v. 30. 

30984) Mwanza, Lake Victoria. 6 vi. 30. 

30985) Ukerewe Id., Lake Victoria. 10. vi. 30. 

Distribution. Also seen on rocks in the river bed at Kilimatinde and 
on rocks on a hillside at Saranda. 

Native names. Milenga (Kisanda,wi) ; finiwinyomo (Kihehe); limvi- 
dunu (Kihehe at Kigogo); nitzubu (Kikinga); luwinzo (Kirungu); 
nsioli (Kifipa). 

Affinities. It was confidently hoped that specimens comparable to 
Mabuya brauni Tornier would be secured in the Ukinga Mountains 
where the holot^pe was collected by Dr. Fiilleborn thirty years ago; 
these expectations were not fulfilled, however, as only typical varia 
were collected there. 

M. brauni was differentiated from varia on three characters: — 

(1) Dorsal scales bicarinate. 

(2) Subocular reaches the mouth but strongly narrowed. 

(3) Color as in varia but lacking the longitudinal stripes. 
The material listed above, shows: — 

(1) Dorsal scales tricarinate, occasionally a few quinquecarinate. 

(2) Great variation is exhibited by the suboculars which broadly 

border the mouth, narrowK- border the mouth, or are en- 
tirely excluded on both sides of the head in one Tandala 



1 { 


M. C. Z. 


5( 


M. C. Z. 


1 


:m. C. Z. 


7 ( 


M. C. Z. 


3 ( 


M. C. Z. 


1 ( 


M. C. Z. 


9< 


:m. C. Z. 


1 ( 


:m. C. Z. 


1 ( 


:m. c. z. 


2( 


:m. c. z. 


9( 


M. C. Z. 


8< 


;m. c. z. 


7 


:m. c. z. 


55 


:m. c. z. 


7 


;m. c. z. 


3 


:m. c. z. 


2 


:m. c. z. 


1 


:m. c. z. 


1 


(M. C. Z. 


1 


(M. C. Z. 


1 


(M. C. Z. 



318 bulletin: museum of comparative zoology 

skink, or on one side only in a Tandala and an Unyanganyi 
specimen. 
(3) During the week spent at Ilolo, some sixty miles from the 
Ukinga Mountains, I saw two varia which, in their uniform 
brown or olive-green color with entire absence of markings, 
were similar to those which I have recorded from the Aber- 
dare Mountains and Northern Guaso Nyiro in Kenya 
Colony. 

One of these is M. C. Z. 30590, the other was too badly damaged to 
be worth preserving. 

It is clear then that brauni can only be distinguished from varia by 
the former's bicarinate scales, it is probably a variant like the one or 
two examples with quinquecarinate scales found in this long series. 
For the present it seems advisable to retain the name as a subspecies 
of varia until more material from the Ukinga Mountains is available. 

Variation. A pair of frontoparietals, subocular usually bordering 
the lip broadly, or very narrowly, or excluded from it entirely in 2% 
of this series; ear lobules generally short, in fact almost indistinguish- 
able in some of the mountain specimens, but long in those from the arid 
thorn-bush country of Unyanganyi and Usandawi though none so long 
as in a paratype of M. varia longiloba Hewitt, which is probably 
synonymous with M. imria damaranus (Peters) 

Coloration. See remarks under affinities. 

Measurements. The largest skink (No. 30951) measures 186 (77+ 
109) mm. 

Breeding. The Ipemi female was gravid on January 7, 1930. 

Diet. Beetles present in the stomach of a Kigogo skink. 

Parasites. Female oxyuroids were removed from an Unyanganyi 
skink. 

Enemies. One was recovered from the stomach of a Lycophidion 
capense uzungioensis at Kigogo and five from as many specimens of 
Trimerorhinus tritaeniatus at Dabaga, Kigogo and Igale. 

Habitat. When rocks are available they are favored by the Variable 
Skink; such was the case at Unyanganyi, Dabaga, Kigogo, Ipemi, 
Mwanza, Kasanga and Bangwe near Ujiji. At the last two places the 
rocks were on the lake shore. At Kigogo, where rocks were very scarce, 
it was interesting to observe the adaptation of this skink to a different 
environment; the few rocks were occupied, but the majority of skinks 
lived in the long grass, basking on dry patches of it and rustling away 
so quickly on one's approach that a glimpse of one was rarely obtained. 
At Igale these skinks were found dwelling in holes in the high earth 



loveridge: African herpetology 319 

bank which constitutes part of the cutting flanking the road. I shot 
the head of one which was protruding from a hole and on withdrawing 
the undamaged specimen found that it was uniformly brow^n and de- 
void of markings, yet in other holes to the right and left of it typical 
varia were secured. I therefore instructed my headman to purchase 
fifty of these skinks at 5 cents each (about 1 cent U. S. currency), the 
total desired were brought in within twenty -four hours but only one 
was similar to the variant which I had secured and it was somewhat of 
an intermediate as it showed faint traces of markings on the sides of 
the head and flanks. Just before leaving Igale I saw a second skink 
entirely devoid of markings. It had been basking on the side of my 
tent and descended to the ground as I approached; it was moving 
away through a tangle of grass when I shot it but was too close so 
that the remains were not worth preserving. Other habitats were 
fallen trees at Dabaga; dry, open forest at Kitungulu, and sandy flats, 
cleared for native gardens, at Kipili. 

Mabuya striata (Peters) 

Tropidolepis7na striatum Peters, 1844, Monatsb. Akad. Wiss. Berlin, p. 36: 

Mozambique. 
Mabuia striata Boulenger, 1887, Cat. Lizards Brit. Mas., 3, p. 204: Zanzibar; 

Zambesi; South Africa; Damaraland. 

1 (M. C. Z. 30986) Bagamoyo. 8. xi. 29. 

4 (M. C. Z. 30987-9) Unyanganyi, Turu. 4. xii. 29. 

2 (M. C. Z. 30990) Tandala, Ukinga Mtns. 11. ii. 30. 
1 (M. C. Z. 30991) Mwaya, Lake Nyasa. 1. iii. 30. 

6 (M. C. Z. 30992) Ilolo, Rungwe. iii. 30. 

1 (M. C. Z. 30993) Tukuyu, Rungwe. 21. iv. 30. 

1 (M. C. Z. 30994) Abercorn, N. Rhodesia. 7. v. 30. 

1 (M. C. Z. 30995) Kitungulu, Urungu. 15. v. 30. 

2 (M. C. Z. 30996) Ujiji, Lake Tanganyika. 28. v. 30. 
1 (M. C. Z. 30997) Shinyanga, Usukuma. 3. vi. 30. 

1 (M. C. Z. 30998) Ukerewe Id., Lake Victoria. 9. vi. 30. 
1 (M. C. Z. 30999) Bukoba, Lake Victoria. 24. vi. 30. 
1 (M. C. Z. 31000) Jinja, Uganda. 30. vi. 30. 
1 (M. C. Z. 31001) Mabira Forest, Lake Victoria. 1. vii. 30. 

Distrihution. Also seen at Dodoma, Saranda and Mwanza. Abun- 
dant on palms and buildings at Bagamoyo; very common at Tukuyu, 
Kasanga and Ujiji; scarce at Unyanganyi where it is replaced by 
M. plajvifrons; scarce at Tandala, Mwaya and Kitungulu, in the last 
locality M. v. varia being ver}" abundant. 



320 bulletin: museum of comparative zoology 

Native name. Mtzuba (Kikinga) ; ulusakand (Kinyakusa for Mabuya). 

Variation. Midbody scale-rows 34-38, several specimens wath 38; 
dorsal scales with 3-5 keels; a pair of frontoparietals. 

Measurements. Largest skink (No. 30990) measures 228 (96+132) 
mm. 

RioPA FERNANDi (Burton) 

Tiliqua fernandi Burton, 1836, Proc. Zool. Soc. London, p. 62: Fernando Po. 
Lygosoma fernandi Boulenger, 1886, Cat. Lizards Brit. Mus., 3, p. 304: Fer- 
nando Po; Nigeria; Cameroon; Gaboon. 

3 (M. C. Z. 31005-7) Entebbe, Uganda. 27. vi. 30. 

Native name. Ngurukisi (Luganda). 

Variaiion. Midbody scale-rows 34. All three agree in every respect 
with Boulenger's redescription except for these additions: — Supra- 
oculars 5-6; owing to the fusion of the 3rd and 4th upper labials, the 
4th and 5th may be below the eye, the 5th labial actually enters the 
orbital ring on one side of the head in Nos. 31005-6, on the other side 
it is narrowly separated. 

Coloration. Boulenger's description, based on preserved material, 
makes no mention of the scarlet sides of this big Gerrhonotus-\\\iQ skink. 

Measurements. The largest specimen (No. 31005) measures 293 
(143+150) mm. 

Diet. Much of the stomach contents was indeterminable but the 
following were found: — A large weevil, two carabid l:)eetles, four hairy 
caterpillars, one millipede, ten slugs. 

Habitat. After sunset on the eve of leaving Entebbe, I was dining 
in the entrance of my tent Avhen one of these skinks came wriggling 
through the grass straight towards me. I captured it and tried to 
bring it back alive but it did not survive the Red Sea. 

I captured the first pair in some soft, sandy soil in the base of a 
rotten stump close to the lake shore. These large lizards are quite 
inoffensive and rely on \'iolent A\Tiggling to effect their escape, they 
are extremely difficult to hold on account of their highly polished 
scales; though their claws are sharp they do not make definite use of 
them as weapons as does a monitor lizard. 

RiOPA SUNDEVALLII SUNDEVALLII (Smith) 

Eumeces (Riopa) sunderallii A. Smith, 1849, Illus. Zool. South Africa, 3, App, 

p. 11: Natal. 
Lyggsotna sundevalli Tornier, 1900, Zool. Jahrb. Syst., 13, p. 599: Discussion of 

East African material. 



loveridge: africax herpetology 321 

Lygosoma sundevallh Schmidt, 1919, Bull. Am. Mus. Nat. HLst., 39, p. 561, 
pi. x.xix: Garamba and Yakululu, Belgian Congo. 

1 (M. C. Z. 31008) Miritini, Kenya Colony. 30. x. 29. 
16 (M. C. Z. 31009-13) Bagamoyo, 9-12. xi. 29. 

1 (M. C. Z. 31027) Kitungulu, Urungu. 15. v. 30. 

2 (M. C. Z. 31028-9) Albertville, Lake Tanganyika. 21. v. 30. 

3 (M. C. Z. 31032-4) Entebbe, Lake Victoria. 27. vi. 30. 

Distribution. Also seen at Changamwe, Kenya Colony. 

Affinities. Schmidt was scarcely correct if by stating that Nieden 
and others united L. modcstum (Giinther) with L. sundevaUii (Smith) 
he meant to imply that the former was a synonym. Superficially this 
would seem to have been their action but in reality both considered it 
a variety, i.e. subspecies, of sundevallii. The Dar es Salaam specimen 
referred to by Tornier was undoubtedly an aberrant, or intermediate, 
sundevaUii and may be matched by one of the Bagamoyo series listed 
above. I have seen a very large skink from Machakos, Kenya Colony 
(Nairobi Museum collection) which has the coloring of sundevaUii but 
has the supranasal fused with the anterior nasal as in modestum. 

The two forms may be distinguished as follows: 

Supranasal not fused with anterior nasal; size larger. 

Back usually much spotted though occasionally 

uniform R. s. sundevallii 

Supranasal fused with anterior nasal; size smaller. 

Back usually uniform browTi, occasionally spotted R. s. modestum 

Variation. Midbody scales smooth (very faint indications of keeling 
in a young Bagamoyo skink) in 26-30 rows (30 on Xo. 31032 only); 
supranasal distinct; nostril usually between two nasals though an in- 
termediate occurs in both the Bagamoyo and Entebbe series where the 
anterior is fused to the supranasal; in an Albertville specimen there 
are three nasals due to a division in one; frontal equals the fronto- 
parietal and parietal together, or is rather longer or shorter; 4th toe 
longer than the 3rd in all these specimens. 

Coloration. Above profusely spotted except the Kitungulu skink 
which is scarcely spotted and looks like modestum. Below uniformly 
white except one Entebbe skink which is well spotted. 

Measuremenis. The largest skink (No. 31008) measures 200 (118+ 
82) mm. 

Diet. Stomach contents of Bagamoyo specimens were : (1) Millipede, 
(2) termites, a small scarab, and wing cases of a larger beetle, (3) a 
long-tailed cricket. 

Parasites. Nematodes were present in a Bagamoyo skink. 



322 bulletin: museum of comparative zoology 

Enemies. The tail of one skink was recovered from the stomach of 
a Rhamphiophis rostratus at Dar es Salaam and a skink from a Lyco- 
phidion capense x acutirostre at Bagamoyo. 

Habitat. Very common in sandy soil beneath rubbish at Miritini; 
abundant under rubbish on the sea coast at Bagamoyo ; beneath logs 
in dry woodland at Kitungulu; beneath bundles of thatching grass at 
Albertville. 

RiOPA suNDEVALLii MODESTUM (Giinther) 

Sepacontias modestus Giinther, 1880, Ann. Mag. Nat. Hist. (5), 6, p. 235: 

Mpwapwa, Ugogo, Tanganyika Territory. 
Lygosoma modestuni Lonnberg, 1907, Reptilia and Batrachia in Sjostedt, 1910, 

Kilimandjaro-Meru Expedition, 1, part 4, p. 8: Kibonoto and Ngare na 

nyuki, Tanganyika Territory. 
Lygosoma ferrandii Loveridge (nee. Boulenger), 1920, Proc. Zool. Soc. London, 

p. 157: Longido and Dodoma, T. T.; 1923, Proc. Zool. Soc. London, pp. 

859 and 962: Nine localities in Central Tanganyika Territory; 1928, Proc. 

U. S. Nat. Mus., 73, Art. 17, p. 66: Dodoma and Mukwese, T. T. 

2 (M. C. Z. 31014-5) Mpwapwa, Ugogo. 23. xi. 29. 

2 (M. C. Z. 31016) Saranda, Ugogo. 30. xi. 29. 

3 (M. C. Z. 31017-8) Unyanganyi, Turn. 4. xii. 29. 

4 (M. C. Z. 31019-22) Handa, Usandawi. 10. xii. 29. 

6 (M. C. Z. 31023-6) Mangasini, Usandawi. 12. xii. 29. 

1 (M. C. Z. 31030) Mw^anza, Usukuma. 6. vi. 30. 

1 (M. C. Z. 31031) Bukoba, Lake Victoria. 24. vi. 30. 

Distribution. Nieden has recorded this form from the Ubena 
Highlands. 

Affinities. In the arid, sandy, thorn-bush steppe of the Central 
Tanganyika plateau is a sandy-brown colored skink which looks 
distinct enough from the large, spotted sundevallii. Some were sub- 
mitted to the British Museum in 1920 and identified as ferrandii, a 
name that I have applied to them ever since. Passing through London 
on my way to East Africa I was afforded the opportunity of examining 
the type oi ferrandii which, as I had begun to suspect, was a different 
creature from the one to which I had been applying the name. In 
accordance with my plans I visited Mpwapwa and was able to secure 
topotypes of modestum. which is the correct name to apply to these 
skinks from the Central plateau. I regard it as a race of sxindevallii 
(of which Parker considers ferrandii a synonym) because of intermedi- 
ates occurring at Machakos, Karungu Bay, Entebbe etc. 

Variation. Based on 27 specimens from 12 localities in Tanganyika. 



loveridge: African herpetology 323 

Midbody scale-rows smooth in 24-28 rows, all but five skinks have 26; 
supranasal is fused with the anterior nasal so that the nostril is be- 
tween the supranasal and a small nasal; frontal usually equals fronto- 
parietal and parietal together but is shorter in six specimens. 

Coloration. Adults are uniformly yellowish brown or dark brown 
above except for the tails which are usually spotted, one adult (No. 
31016) has retained the spotting of the juveniles which, in addition to 
the dorsal spotting sometimes possess a dark lateral band. Ten miles 
east of Unyanganyi a unique specimen (No. 31018) was taken under 
a log lying on red soil, this skink was uniformly red beneath and the 
color is retained even after two years' immersion in formalin and 
alcohol. 

Measurements. The largest skink (No. 18689) is from Ikikuyu and 
measures 180 (95+85) mm., but this is an exceptionally large in- 
dividual, and much above the average. 

Enemies. At Mangasini one skink was recovered from the stomach 
of a Coronella semiornata and two from Psammophis biseriatus. 

Habitat. The Mpwapwa skinks were taken in sandy soil under a 
rotting tree stump in the bank of a dry stream-bed. Saranda and 
Handa specimens were all found beneath logs. At Unyanganyi 
among dead leaves at the base of a bush. They were most numerous at 
Mangasini where at least a dozen were seen in the course of a twenty 
minutes' walk over a kopje after sunset, at which time they emerge 
from their retreats and WTiggle about. 

Ablepharus boutonii africanus Sternfeld 

Ablepharus peronii Peters {nee. Cocteau), 1854, Verh. Akad. Wiss. Berlin, p. 

619: Cabageira, Mozambique. 
Ablepharus boutoni africanus Sternfeld, 1918, Abh. Senckenb. Nat. Ges., 36, p. 

423: Manda Island, Malindi and Pemba Island. 
Cryptoblepharus boutonii peronii Loveridge (nee. Cocteau), 1929, U. S. Nat. 

Mus. Bull. 151, p. 80: discussion on races. 

26 (M. C. Z. 31036-45) Dar es Salaam. 6. xi. 29. 

Affinities. Mertens has recently (1931, Zool. Jahrb. Syst., 61, pp. 
63-210) made a very thorough revision of all the races of boutonii and 
shown that Cryptoblepharus cannot be retained as a distinct genus. 
As he considers africanus a valid race I accept his decision. 

Variation. Midbody scale-rows 22-24, with an average of 23 for the 
whole series. 

Measuremenis. The largest skink measures 114 (42+72) mm., but 



324 bulletin: museum of comparative zoology 

is exceeded in body length by one of 107 (48+59) mm., both are 
sh'ghtly surpassed by some which I recorded in 1920. 

Breeding. Most, or all, of the females have the ovules enlarged and 
measuring up to 10 x 5 mm. 

Ablepharus wahlbergii (Smith) 

Cryptoblepharus wahlbergii Smith, 1849, Illus. Zool. S. Africa, 3, App., p. 10: 

Natal. 
Ablepharus wahlbergii Loveridge, 1929, U. S. Nat. Mus. Bull. No. 151, p. 79: 

Nairobi and Mt. Kenya, Kenya Colony. 

15 (M. C. Z. 31046-54) Bagamoyo. 9-12. xi. 29. 
3 (M. C. Z. 31055-7) Mpwapwa, Ugogo. 23. xi. 29. 
3 (M. C. Z. 31058-60) Masiliwa, Turu. 10. xii. 29. 
3 (M. C. Z. 31061-3) Kitungulu, Urungu. 15. v. 30. 
1 (M. C. Z. 31064) Nyamkolo, Lake Tanganyika. 9. v. 30. 

Distrihuiion. Seen also at Mangasini and Kasanga. 

]'ariation. Midbody scale-rows 22-28, this astonishing range was 
to be found in the Bagamoyo series alone as well as elsewhere, the 
average of twenty-one skinks is 25 scale-rows ; the normal number of 
3 supraoculars is present in every skink. 

The Nyamkolo specimen, which also has 28 scale-rows, looks so 
different from any wahlbergii that I have ever seen that I felt confident 
in the field that it represented a different species. Examination in the 
laboratory, however, shows that the head shields and all other charac- 
ters are normal so that it must be its large girth and strange coloring 
that produce the illusion. 

Coloraiion. An Mpwapwa skink had the throat white but the rest 
of the lower surface salmon-red. The Nyamkolo skink had the throat 
spotted. 

Measurements. The Nyamkolo skink measures 83 (50+33) mm., 
its tail being regenerated; the next largest specimen (No. 31049) 
measures 94 (45+49) mm. 

Breeding. At Bagamoyo the females held large ova, at Mpwapwa 
two eggs and a young one were unearthed among the rotting roots of a 
large tree-stump in the sandy soil of a dry stream-bed. Two of these 
eggs hatched as I picked them up and young ones, measuring 34 
(17+17) mm., wriggled out. I had always supposed the species to be 
viviparous like the majority of skinks. At Masiliwa two very small 
young, perhaps a few weeks old, were taken after sunset as they were 
running about in a drift of dry leaves at the base of a ])ush. 



loveridge: African herpetology 325 

Enemies. At Bagamoyo two and four tails were recovered from the 
stomachs of Lycophidion capen.se >< acutirostre and Phihthamnus s. 
semivariegatus respectively. 

Habitat. Very abundant among grass roots on sandy soil at Baga- 
moyo; among fallen leaves under a mango tree at Mpwapwa; an adult 
under a log at INIasiliwa; on rocky kopje at Mangasini; both in dry 
woodland and among dry leaves in a remnant of rain forest at Kitun- 
gulu. It was apparently this species which was several times seen 
running over rocks near the water's edge of Lake Tanganyika at 
Kasanga and elsewhere. 

Ablepharus megalurus Nieden 

Ablepharus megalurus Nieden, 1913, Mitt. Zool. Mus. Berlin, 7, p. 89: Kin- 
janganja in Turu, 4° 50's., i.e. Unyanganyi, Turu, Tanganyika Territory. 

8 (M. C. Z. 31065-9) Mangasini, Usandawi. 14. xii. 29. 
5 (M. C. Z. 31070-4) Saranda, Ugogo. 18. xii. 29. 

Distribution. Hitherto only known from the holotype, Saranda is 
exactly fifty miles due south of Unyanganyi, Mangasini lies betw^een 
the two but about fifteen miles east. 

Variation. Midbody scale-rows 20-22, apart from this the series 
agrees very closely with the original description except in one par- 
ticular which may be translated thus, "Through their entire length 
the paired frontoparietals separate from the interparietal," I imagine 
that this was intended to read, "Through their entire length the paired 
frontoparietals separate the frontal from the interparietal" for this 
is the arrangement in my specimens which are megalurus beyond any 
shadow of doubt. 

Coloration in life. Naturally this differs somewhat from Xieden's 
description based on an alcoholic specimen though the markings cor- 
respond. Above, bronze, from each eye or nuchal scale, a light silver 
line edged above and below by black, extends to the root of the tail; 
owing to the centre of each scale being darker in some individuals 
such specimens have the appearance of dusky lateral stripes corre- 
sponding to the scale-rows; tail vivid pale blue, more or less longitudi- 
nally striped. Below, satiny-white or tinged with blue; throats of 
three of the larger specimens darker. 

Measurements. The largest skink measures 92 (32+60) mm., and 
the smallest 49 (20+29) mm., both from Mangasini. This is un- 
doubtedly the smallest species of lizard inhabiting East Africa. 

Habitat. When on the way to Unyanganyi for the sole purpose of 



326 bulletin: museum of comparative zoology 

securing a topotypic series of this skink, we were delayed at Saranda 
several days. The country was very desiccated and, on November 
28th, ha^•ing just shot an owl after sunset, I was hurnnng back to the 
road with a gun in one hand and the owl in the other. Salimu came to 
meet me just as I was about to jump a shallow trench, at the same 
instant I saw a skink scuttle over the edge of the trench and disappear 
into one of the numerous fissures of the sun-baked black cotton soil. 
I recognised it as a species new to me but, though we returned to the 
place at all hours of the day and Salimu searched the trench for a 
hundred yards in either direction we never saw the reptile again. 

At Unyanganyi, this time the sun was low but had not quite dis- 
appeared, I saw a second specimen which vanished into a mere slit of 
similar soil at the base of a bush in a desiccated irhiujwe. Next day, 
despite the almost iron-like hardness of the ground we dug up the 
whole vicinity to the depth of a foot without meeting with any success. 

Shortly after we reached Mangasini the long-delayed rains broke 
and the baked plains below the camp began to flood. Salimu returned 
with the first pair of these skinks and by showing them to all the native 
children that visited camp and offering the high price of twelve cents 
(U. S. currency) six more were secured. It is next to impossible for a 
white man to capture these elusi^•e little creatures by reason of their 
unusual habitat. They live in the fissures of the cracked soil, such 
fissures often being six feet or more in depth; they usually bask on the 
sides of the fissures (by which I mean within the crack) in the early 
morning or late afternoon at which times the sun strikes obliquely. 
So wary are these reptiles that they run down the fissure as soon as a 
shadow falls upon them and by stalking I could never get within six 
feet. 

At Saranda, Salimu, assisted by Abedi, spent the greater part of a 
day hunting megalura and only succeeded in capturing five. These 
were got by WTiggling up to the point above where the skink had been 
seen and dabbing a wad of soft material over the reptile, the edge of 
the cloth was then turned back with care until the specimen was 
located and its head or neck grasped with a forceps, even so a large 
proportion parted with their tails. 



Melanoseps ater (Giinther) 

Herpetosaura atra Gunther, 1873, Ann. Mag. Nat. Hist. (4), 12, p. 147: Zam- 
besi; Peters, 1882, Reise nach Mossamb., 3, p. 81. 
Melanoseps ater Boulenger, 1887, Cat. Lizards Brit. Mus., 3, p. 422. 



loveridge: African herpetology 



327 



Melanoseps ater var. longicauda Tornier, 1900, Zool. Jahrb. Syst., 13, p. 602; 

Masailand & Korogwe, Tanganyika Territory; Loveridge, 1923, Proc. 

Zool. Soc. London, p. 963: Mkata Station, Tanganyika Territory. 
Melanoseps ater longicauda Barbour & Loveridge, 1928, Mem. Mus. Comp. 

Zool., 50, p. 169: Vituri, LTluguru Mtns., Tanganyika Territory. 

1 (M. C. Z. 31075) Mpwapwa, Ugogo. 22. xi. 29. 

2 (M. C. Z. 31076-7) Kigogo, Uzungwe Mtns. 23. i. 30. 

Affinities. This additional material confirms the opinion expressed 
in 1928 that Tornier's longicauda was not specifically distinct from 
ater. In all probability the type of ater was a female and that of 
longicauda was a male, hence the longer tail.^ Arranged from south 
to north the available data is as follows: — 









Length 






Mid- 








of 


Length 


Tail 


body 








head and 


of 


into 


scale- 




Locality 


Sex 


body 


tail 


H &B 


rows 


Type of ater 


Zambesi River 


? 9 


160 mm. 


43 mm. 


3.72 


22 


M. C. Z. 31076 


Kigogo 


9 


108 " 


28 " 


3.85 


26 


" 31077 


(( 


9 


210 " 


- 


- 


28 


" 31075 


Mpwapwa 


9 


87 " 


28 mm. 


3.10 


20 


18356 


Mkata Station 


9 


82 " 


30 " 


2.73 


18 


" 24235 


Vituri 


d^ 


124 " 


- 


- 


22 


Cotjrpes of \ 


Korogwe 


? 


52 " 


- 


- 


19 


longicauda / 


Masailand 


? c? 


71 " 


41 mm. 


1.73 


19 



Coloration in life. Based on the larger Kigogo female. Above, and 
on sides, uniformly steely-blue-black. Below, white, each scale with 
a black centre thus giving the appearance of about a dozen longitudi- 
nal lines from throat to anus but on the tail forming a diamond-shaped 
pattern. 

Breeding. Ovules small. 

Diet. A large caterpillar in the stomach of one Kigogo skink. 

Habitat. The Mpwapwa specimen was taken in sandy debris among 
the rotten roots of a fallen tree fifty feet from a stream, though the 
actual site was bone-dry. The Kigogo skinks were taken by natives 
engaged in clearing the vegetation from black alluvial soil on a grass- 
grown hillside at the very edge of the temperate rain forest. 

1 I asked Mr. H. W. Parker for further information on this point and after examining the type, 
he replied: " It has not been dissected but is apptirently a female: the tail looks as if it may have 
been injured, but I should not like to say definitely that it had: there are one or two minor 
irregularities of the scaling just where it rounds off for the tip: there is no terminal scute, but 
this looks like a postmortem loss and not due to an injury in life. 



328 bulletin: museum of comparative zoology 

AXELYTROPIDAE 

Feylinia currori elegans (Hallowell) 

Acontias elegans Hallowell, 1852, Proc. Acad. Nat. Sci. Philad., p. 64: Liberia 
(? Gaboon, fide K. P. Schmidt). 

1 (M. C. Z. 31078) Entebbe, Lake Victoria. 27. vi. 30. 

Distribution. Feylinia currori (whether elegans or the t^'pical form 
I cannot say) has been reported by Boulenger from the Sesse Islands 
in Lake Victoria and by Xieden from Bukoba on the western shore of 
the lake; the present record appears to be the most easterly published 
but there are three Entebbe examples collected by Sir Harry Johnston 
(1900) and Hoare (1929) in the British Museum collection. 

Affinities. F. elegans was tentatively referred to the synonymy of 
F. currori by Boulenger in 1896 (Cat. Lizards Brit. Mus. 3, p. 431) but 
in 1919, Schmidt (Bull. Am. Mus. Nat. Hist. 39, p. 605) after studying 
the type and other examples, proposed recognising elegans as a species 
which differs from currori in having the ocular in contact with the 
second labial (third in currori) and cut off from the third by a post- 
ocular. 

Mr. H. W. Parker has kindly sent me the following information 
based on fifteen specimens in the British Museum. Of these, four are 
from Entebbe and Msori, Uganda and agree in having the eye in 
contact with the second labial which, he remarks, appears to cor- 
respond to a fusion of the first and second in the others ; eleven West 
Coast (Nigeria, Cameroon, French Congo, Gaboon and Angola) 
specimens agree with the two currori (Cameroon and Belgian Congo) 
in the Museum of Comparative Zoology in having the eye in contact 
with the third labial. The distribution is obviously peculiar if the 
type locality of elegans is correct but a later paper by Hallowell raises 
doubts as Schmidt points out. As our Entebbe specimen agrees with 
the description of elegans I use the name tentatively pending increase 
in our knowledge of variation within this genus. 

Variation. Scale-rows behind head 24, midbody 26, immediately in 
front of anus 23. 

Measurements. Total length 237 (206+31) mm. 



loveridge: African herpetology 329 



C H AM AELEOXTID AE 

In 1887, at the time of the publication of Boulenger's third volume 
of the Catalogue of Lizards in the British Museum, he was able to 
define three genera of chameleons (p. 438) as follows: — 

Claws simple, scales on soles smooth ; tail at least as 
long as the body Chamaeleo7i 

Claws simple; scales on soles spinose; tail shorter 
than the body Brookesia 

Claws bicuspid; scales on soles spinose; tail shorter 

than the body Rhamjjholeon 

In his redescription of Chamaeleon anchietae, however, he states, 
"Tail slightly longer than head and body." though this contradicts 
Bocage's measurements of the type series of five specimens. That 
these were correct is apparent from the plate of anchietae which ap- 
peared some years later in Bocage's work on the herpetology of the 
Congo and Angola. In an example of anchietae obtained during the 
present expedition the tail is considerably shorter than the length from 
snout to anus. 

During the interval that has elapsed since 1887 a large number of 
reptiles variously described as Brookesia or Rhampholeon have been 
discovered and it is obvious that the genus Rhampholeon, proposed by 
Gunther in 1874, will have to be placed in the synonymy of Brookesia, 
Gray 1864. My reasons for reaching this conclusion are as follows: 

In 1893 Matschie described Chamaeleon (Brookesia) temporalis from 
Derema, at the same time he described Chamaeleon (Brookesia) 
brevicaudatus from the same type locality. Both have simple claics but 
because temporalis had smooth soles Tornier left it in the genus 
Chamaeleon but transferred brevicaudatus to Rhampholeon because of 
its very spinose soles. Yet so closely related are the two that I mis- 
identified a temporalis, obtained at Amani in 1926, for brevicaudatus. 
This would not have happened had not Werner, following Tornier, left 
temporalis in Chamaeleon when he revised the family and despite the 
excellent figures given by Tornier which show its stumpy tail (vide. 
Tornier, 1897, Die Kriechthiere Deutsch-Ost-Afrikas, pi. ii, figs. 5 & 7). 

Angel, in his revision of the Malagasy Brookesia in 1929, points out 
that a number of species have spinose soles but at least three have 
smooth soles. So that smooth and spinose soled forms occur both 
among the Malagasy and Continental forms, and though bicuspid 
claws are found only among Continental reptiles, simple claws occur 
in both Malagasy and Continental species. If we arbitrarily continue 



330 bulletin: museum of comparative zoology 

to define genera on the basis of simple or bicuspid claws it divides 
closely related species such as temporalis and pJutyccps also brcvicauda- 
tus and hrachyurus which form parallel groups of simple and bicuspid 
claws, the former having a rostral process in common, the latter with- 
out any such process. The following key will cover the suggested 
arrangement. 

Tail prehensile, usually as long as, or longer than, the 

body ; soles smooth; claws simple Chamaeleon 

Tail not prehensile, always shorter than the body; 

soles smooth or spinose ; claws simple or bicuspid . . . Brookesia 



Chamaeleon gracilis gracilis Hallowell 

Chamaeleo gracilis Hallowell, 1842, Journ. Acad. Nat. Sci. Phila., p. 324, pi. xviii: 
Monrovia, Liberia. 

3 (M. C. Z. 31081-2) Entebbe, Uganda. 27. vi. 30. 

Measurements. Largest female (No. 31082) measures 282 (147+ 
135) mm. The tails are included 0.48 times in the total length. 



Chamaeleon dilepis roperi Boulenger 

Chamaeleon roperi Boulenger, 1890, Proc. Zool. Soc. London, p. 85, pi. viii, fig. 4: 
Kilifi, north of Mombasa, Kenya Colony. 

cf 9 (M. C. Z. 31083-4) Mainland near Mombasa. 29. x. 29. 
c? (M. C. Z. 31085) Changamwe, Kenya Colony. 31. x. 29. 

Measuremenis. Larger male measures 183 (89+94) mm. The tails 
of the males are included 0.47-to 0.51 times in the total length, that 
of the female 0.48 times. 

Diet. The stomachs of the Kilindini specimens were examined and 
that of the male found to hold many greenbottle flies, while in the 
female there was the hind leg of an orthopteran and numerous remains 
of beetles among which weevils and a buprestid were recognizable. 

Habitat. At Kilindini the female was taken as she stalked across 
some open grassland, the male was in a mimosa bush. Curiousl.y 
enough I pulled off the larval case of a psychid moth from a branch 
close beside the chameleon without seeing the reptile so closely did it 
match its surroundings, it was then pointed out V)y the native who 
accompanied me. The Changamwe male was taken as it ascended a 
tree trunk. 



LOVERIDGE: AFRICAN HERPETOLOGY 331 



Chamaeleon dilepis quilensis Bocage 

Chamaeleo dilepis var. quilensis Bocage, 1866, Join. Sci. Math. Phys. Nat. 
Lisboa, 1, p. 59: Rio Quillo, Angola. 

1 (M. C. Z. 31086) Iringa, Uhehe. 30. i. 30. 
7 (M. C. Z. 31087-93) Matema, near Mwaya. 28. ii. 30. 
50 (M. C. Z. 31094-110) Mwaya, Lake Nyasa. 1-8. iii. 30. 
Eggs and 38 (M. C. Z. 31111-35) Ilolo, Rungwe. 15. iii. 30. 

4 (M. C. Z. 31136-9) Tukuyu, Rungwe. 13. iii. 30. 

4 (M. C. Z. 31140-3) Nyamkolo, Lake Tanganyika. 9. v. 30. 

1 (M. C. Z. 31144) Ujiji, Lake Tanganyika. 29. v. 30. 

1 (M. C. Z. 31145) Mwanza, Lake Victoria. 6. vi. 30. 

Distribution. Also seen at Mwandemeres. This race has already 
been recorded from Iringa under the name of C. parvilobus Blgr. In 
his 1913 paper on the reptiles of German East Africa, Nieden did not 
list this form. It is very doubtful if it is a true geographical race. 

Native names. Ulmoifi (Kinyakusa); luvivu (Kirungu). 

Variation. The tails of 40 males range from .46 to .53 of the total 
length, with an average of .49; those of the 57 females are from .44 
to .53, also with an average of .49; in 11 young the average is .47. 

Coloration. The following striking variations were noted at Ilolo: 
(i) A very dark green spotted all over with yellow and pale green; on 
the tail, bands of dark and pale green alternate, lateral stripe absent. 
(ii) Pale yellow green more or less finely mottled with yellow; a white 
stripe bordering the buccal opening covers both upper and lower 
labials, another on the flank but not extending as far as the hind limb ; 
the foot margined with white one scale in width, (iii) cf . Ashy grey 
with darker bands on tail and sides spotted with blue-black, (iv) 
Dark olive with lips and lateral stripe of china-white also a spot of 
white near the occipital lobe and another on the middle of the side. 
(v) A specimen which was wrathful or scared. Absolutely black ex- 
cept for the light ventral line and the interstitial gular skin which was 
orange. 

Measuremenis. The largest male measures 265 (124+141) mm.; 
the largest female 286 (133+153) mm.; the smallest specimen only 
66 (34+32) mm. is from Mwa3'^a. 

Breeding. Of 23 females taken at Ilolo, all except the largest, which 
was presumably sterile, carried well-developed eggs. The numbers in 
ten females examined ranged from 26 to 41 with an average of 31. 
Their measurements were as follows : 



332 



bulletin: museum of comparative zoology 



(1) 25 measuring 8.5 mm. diameter. 

(2) 35 " 9 mm. 

(3) 38 " 9 mm. 

(4) 35 " 12 X 6.5 mm. 

(5) 34 " 12 X 7.5 mm. 



(6) 41 measuring 13 x 7.5 mm. 

(7) 30 " 14 X 8 mm. 

(8) 30 " 14x8 mm. 

(9) 24 " 15 X 7 mm. 
(10)25 " 16x7 mm. 



Diet. The distinguishable contents of the stomachs of these ten 
females was : (i) Beetles, two snails, (ii) two beetles, three snails, (iii) 
beetle, two grasshoppers, snail, (iv) beetle, grasshopper, hawkmoth 
larva, snail, (v) beetles, grasshoppers, (vi) many beetles, many grass- 
hoppers, two greenbottle flies, (vii) several large grasshoppers, (viii) 
small grasshopper, big green caterpillar, (ix) many small grasshoppers, 
a butterfly or moth, (x) many greenbottle flies, butterfly, snail. 



Chamaeleox dilepis dilepis Leach 

Chamaeleo dilepis Leach, 1819, in Bowdich, Miss. Ashantee, App. p. 493: 
Gaboon. 

1 (M. C. Z. 31146) Kilimatinde, Ugogo. 27. xi. 29. 

1 (M. C. Z. 31147) Unyanganyi, Turu. 7. xii. 29. 

8 (M. C. Z. 31148-55) Mangasini, Usandawi. 12-16. xii. 29. 

1 (M. C. Z. 31156) Kikuyu, Ugogo. 23. xii. 29. 

4 (M. C. Z. 31157-60) Ukerewe Id., Lake Victoria. 11. vi. 30. 

Distribution. Already recorded by Xieden from Kilimatinde and 
Mwanza. 

Native names. Wambu (Kinyaturu); kathlange (Kisandawi); luivu 
(Chigogo). 

Variation. The tails of the 7 males range from 0.45 to 0.52 of the 
total length, with an average of 0.47S; those of the 13 females are from 
0.46 to 0.52, with an average of 0.496. 

Measurements. The largest male measures 283 (137+146) mm.; the 
largest female 299 (150+149) mm. Both from Ukerewe Island. 

Diet. Stomach contents were as follows : — (1) Full of beetle remains, 
(2) distended with coccinelids and other beetles, grasshoppers and a 
large field cockroach, (3) winged termites, and its owti cast slough, 

(4) winged termites, a common striped moth, its own cast slough, 

(5) termites, large locust, (6) beetles, green elytra of rose beetle, grass- 
hoppers. A seventh chameleon disgorged winged termites and a large 
grasshopper. 

Enemies. One chameleon was recovered from the stomach of a 
Boomslang (Displiolidu^ typu^) at Mangasini. 

Habitat. A rather emaciated male, though its stomach was full of 



loveridge: African herpetology 333 

beetle remains, was taken on a manyara hedge at Kilimatinde. As 
so many creatures shun the poisonous manyara it seems worth re- 
cording. 

Chamaeleox bitaeniatus bitaeniatus Fischer 

Chamaeleo bitaeniatus Fischer, 1884, Jahrb. Hamb. Wiss. Anst., 1, p. 23, pi. ii, 
fig. 7: Masailand, East Africa. 

59 (M. C. Z. 31161-83) Entebbe, Uganda. 27. vi. 30. 
2 (M. C. Z. 31184-5) Mabira Forest, Uganda. 1. vii. 30. 

Variation. Only 25 adults were selected for measurement, of these 
the tails of 11 males range from 0.44 to 0.51 of the total length, with 
an average of 0.469; those of 14 females are from 0.42 to 0.48, with an 
average of 0.462. 

Measurements. The largest male measures 159 (81 + 78) mm.; the 
largest female 157 (81 + 76) mm.; the smallest example 49 (28+21) 
mm. 

Breeding. ]Many of the females held embryos. 

Chamaeleon bitaeniatus hohnelii Steindachner 

Chamaeleon hohnelii Steindachner, 1891, Sitzber. Akad. Wiss. Wien, 100, p. 
309, pi. 1, fig. 2: Leikipia, Kenya Colony. 

d' 9 (M. C. Z. 31384-5) Kabete, Kenya Colony. 7. vii. 30. 

Variation. Both variation and measurements are within the range 
shown by the 431 specimens in the United States National Museum 
(Loveridge, 1929, U. S. Xat. Mus. Bull. Xo. 151, pp. 87-89). 

Breeding. The female holds very large ova. 

Chamaeleon anchietae Bocage 

Chamaeleo anchietae Bocage, 1872, Jorn, Sci. Lisboa, p. 72, fig.: HuHla, Mossa- 

medes, Angola. 
Chamaeleon anchietae Boulenger, 1887, Cat. Lizards Brit. Mus., 3, p. 452. 

9 (M. C. Z. 31186) Panga Mawe, Uzungwe Mtns. 8. i. 30. 

Distribution. This constitutes the first record of this scarce species 
in East Africa, and appears to be the only specimen collected since the 
type series was described sixty years ago. 

Variation. Though he had no specimen Boulenger has added to the 
original description, "Tail slightly longer than head and body." This 



334 bulletin: museum of comparative zoology 

contradicts the measurements given by the author and is opposed to 
his excellent figure which later appeared in the Herpetology of Angola 
and the Congo. In my specimen the tail is 0.36 of the total length. 

Measuremenfs. Total length 113 (72+41) mm. 

Breeding. Ova are small and undeveloped. 

Diet. Stomach empty. 

Parosites. A nematode (Sfrongyluris sp.) was present in the intestine. 

Habitat. My attention was attracted by the very brilliant green 
coloring of this chameleon which was walking through short, freshly- 
springing grass on a recently burnt-over hillside. It would seem prob- 
able that the short tail indicates a species which has become adapted 
to life on low shrubby plants on wind-swept mountainsides. 



Chamaeleon goetzei Tornier 

Chamaeleon goetzei Tornier, 1899, Zool. Anz., 22, p. 413, fig. 3: Uhehe, Tan- 
ganyika Territory. 

2 (M. C. Z. 31187-8) Dabaga, Uzungwe Mtns. 1. i. 30. 

1 (M. C. Z. 31189) Lukungu, Ubena Mtns. 8. ii. 30. 

2 (M. C. Z. 31190-1) Ihenye, Ukinga Mtns. 8. ii. 30. 

4 (M. C. Z. 31192-5) Mangoto, Ukinga Mtns. 10. ii. 30. 

2 (M. C. Z. 31196-7) Tandala, Ukinga Mtns. 11. ii. 30. 

3 (M. C. Z. 31198-200) Bulongwa, Ukinga Mtns. 12. ii. 30. 

3 (M. C. Z. 31201-3) Madehani, Ukinga Mtns. 14. ii. 30. 
60 (M. C. Z. 31204-20) Ilolo, Rungwe district. 15-30. iii. 30. 

4 (M. C. Z. 31232-5) Igale, Poroto Mtns. 30. iv. 30. 

Distribution. Also seen at Lukowa in the Ubena Mountains. Listed 
by Nieden in 1913 as occurring in Ubena, Uhehe and at Tukuyu. 

Native name. Tanatzi (Kikinga). 

Variation. The tails of 42 males range from .47 to .54 of the total 
length, with an average of .51; those of 32 females are from .50 to .54 
with an average of .52. 

Coloration in life. 9 Madehani. Above and sides of head, limbs, 
flanks and tail, olive brown; vertebral crest rich dark green; four large 
patches of green on each side, their upper portions interrupting a light 
buff line which extends from the eye to the base of the tail; a second 
buff line extends only from the eye to the fore arm and a third buft' 
line from the posterior portion of the upper lip to axillary region being 
immediately above the concealed black gular streak; a little black, 
almost as an interrupted continuation of the black gular streak, shows 
here and there along the lowest part of the flank when the reptile is 



loveridge: African herpetology 335 

alarmed or annoyed. Below the throat is lemon-yellow except for two 
longitudinal, white streaks which merge into the narrow ventral area. 

The first specimen found at Dabaga, when chloroformed, became 
entirely plumbeus except for the lips which remained white. Tlu-ee 
females from Bulongwa were a vivid green in life. 

Measurements. The largest of 42 males measures 208 (95+113) 
mm.; the largest of 32 females is 190 (89+101) mm.; both from Ilolo. 

Breeding. Young chameleons of 48 (25+23) mm., 51 (26+25) mm., 
and 52 (28+24) mm. were taken at Ihenye, Ilolo and Nyamwanga on 
the dates given above. 

Diet. The stomach contents of fifteen chameleons from Ilolo and 
Nyamwanga were as follows: (i) House fly and many bugs, (ii) beetles, 
two large ants, three caterpillars, (iii) beetle, large moth, two cater- 
pillars, spider, (iv) tortoise beetle, bug, large dipteron, grasshopper, 
(v) grasshopper, cockroach, (vi) grasshopper, three caterpillars, (vii) 
beetle larva, ant, bluebottle, (viii) beetles and many other masticated 
insects, (ix) beetles, froghopper, (x) beetles, bug, (xi) beetle, bug, 
(xii) beetles, ants, (xiii) beetles, caterpillar, (xiv) beetles, three cater- 
pillars, grasshopper, (xv) beetles, caterpillar, grasshopper, froghopper, 
ant, spider. 

Parasites. Nematodes {Strongyluris ? hrcvicaudaia) were present. 

Enemies. Both at Lukowa and Lukungu in the Ubena IMountains I 
came upon dead and decaying chameleons in the path. Wakinga 
natives told me that they always killed chameleons "because they spit 
venom"; a native interpretation of the rapidly projected tongue. 

Defence. When taken up the first chameleon hissed and ichistled, 
the whistle always appears to follow the hiss and I have never known 
chameleons of any other species to make this sound. Later I was able 
to confirm the observation and found that whistling was a common 
accompaniment of the creature's protestations when seized the first 
time but that it is not repeated after the reptile has been handled, so 
that it is unlikely to be noticed in specimens which may reach zoologi- 
cal gardens. This unique achievement for a chameleon is doubtless 
connected with the peculiar large black patches concealed in folds on 
either side of the throat and looking almost exactly like the singing 
pouches of certain male frogs of the genus Rana. 



Chamaeleon tempeli Tornier 

Chamaeleon tempeli Tornier, 1899, Zool. Anz., 20, p. 411, fig. 2: Utschungwe 
(i.e. Uzungwe) Mountains in Uhehe, Tanganyika Territory. 



336 



bulletin: museum of comparative zoology 



Chamaeleon tempeli var. wolfi Tornier, 1900, Zool. Jahrb. Syst., 13, p. 614, fig. 
G (note, the figures are reversed but the error was corrected in separates in 
Tornier's own hand, also the fig. is labelled wolffi, not wolfi): Tardalla (i.e. 
Tandala, Ukinga Mountains, Tanganyika Territory). 

22 (M. C. Z. 31236-48) Dabaga, Uzungwe Mtns. 1. i. 30. 

2 (M. C. Z. 31249-50) Mufindi East, Uzungwe Mtns. 9. i. .30. 
24 (M. C. Z. 31251-60) Kigogo, Uzungwe Mtns. 13. i. 30. 

1 (M. C. Z. 31261) Mangoto, Ubena Mtns. 10. ii. 30. 
14 (M. C. Z. 31262-5) Tandala, Ukinga Mtns. 11. ii. 30. 

1 (M. C. Z. 31266) Bulongwa, Ukinga Mtns. 12. ii. 30. 
51 (M. C. Z. 31267-85) Madehani, Ukinga Mtns. 13-28. ii. 30. 

Distribution. According to Xieden, in 1913 typical tempeli was only 
known at that time from the Uzungwe Mountains and Ufipa to Lake 
Tanganyika while of the subspecies wolfi only the type was known. 

Native names. Lumwilifivi (Kihehe for all chameleons); tanatzi 
(Kikinga, but not specific). 

Affinities. C. t. wolfi was described from a single specimen which 
had the two anterior projecting rostral scales of C. t. tempeli fused into 
a single scale or tiny horn. ^Yith the object of finding whether this 
character was constant I camped at Tandala for one night and secured 
fourteen topotypes. Of these only two, a cf and 9 , have the terminal 
rostral scales fused in this fashion, the remaining twelve have the 
normal arrangement of the typical form from the Uzungwe Mountains. 
At Madehani forty-five are of the tempeli type and two of the wolfi, 
three others are somewhat intermediate. The ]Mangoto and Bulongwa 
chameleons which should agree with wolfi are also of the tempeli type. 
If further proof is required that wolfi is anything but a variant it might 
be remarked that there are five chameleons of the wolfi type in the 
topotypic series of tempeli from Kigogo and Dabaga. 

Variation. The tails of 31 males range from .49 to .53 of the total 
length, with an average of .50; those of 24 females have the same range 
but average .53. 

Coloration in life. At Dabaga it was noted that for a chameleon the 
coloring was most unusual in shades of ochre, buff, olive, reddish- 
brown, blue grey, black and white. The head of ochre; the points of 
the spines are tipped with black; there are reddish-brown blotches 
alternative with blue grey along the spine; the sides are blotched with 
olive and blue grey, the limbs are tinged with yelloAv; underparts 
whitish; the interstitial gular skin black. Another specimen was 
colored in shades of rotten wood — browns and black. These shades 
are characteristic of a common shrub {Protea sp.) which is abundant 
at Dabaga. 



loveridge: African herpetology 337 

Of the two taken at Mufindi one showed the dorsal line and spines 
all black and three black marks on each flank. In the other the top 
of the head was a pinkish-ochre; just below the dorsal spines was a 
light lateral streak while the intermediate area was raw sienna; the 
throat was chalky -white with black interstitial skin. 

A young one at Kigogo, taken on green leaves, was distinctly 
greenish in shade; it is quite usual for one side of these chameleons to 
be in paler tints than the other. I caught a male in the act of descend- 
ing a tree trunk the lichen of whose bark he matched to perfection. 

Unfortunately no notes were taken of Ukinga specimens but my 
first impression was that the series from this region differed somewhat 
in general color, being less ochraceus and more olivaceus than those 
from the UzungAve Mountains. 

Measuremenis. The largest male measures 213 (109+104) mm.; 
the largest female 203 (98+105) mm. 

Breeding. Both at Kigogo and Madehani young specimens were 
collected which measured 61 (31+30) mm. 

Did. The distinguishable contents of the stomachs of Kigogo 
specimens were: (i) Beetles, caterpillars, flies, spider, (ii) beetles, 
caterpillars, flies, (iii) beetles, caterpillars, flies, (iv) beetles, flies, 
spider, (v) beetles, caterpillars, spider, (vi) beetles, two caterpillars, 
several spiders, (vii) beetles etc. (viii) beetles, flies, (ix) beetles, (x) 
beetles, (xi) beetles, (xii) caterpillars. 

At Mangoto and Madehani the following were recognizable: (i) 
beetles, caterpillar, (ii) beetles, caterpillar, (iii) beetles, caterpillar, 
(iv) beetles, caterpillars, (v) beetles, caterpillars, (vi) beetles, cater- 
pillar, grasshoppers, (vii) beetles, caterpillar, spider, (viii) caterpillar, 
flies, bug, (ix) beetles, (x) flies, grasshoppers. 

Parasites. Nematodes {Strojigyluris brevicaudata) appear to be in- 
variably present in the intestines of this species and were collected at 
Dabaga, Kigogo, Mangoto, Tandala and Madehani; cestodes {Oocho- 
ristica theileri) were also found. 

Enemies. At Dabaga one was recovered from the stomach of a hawk 
{Astur t. sparsimfasciatus). 

Habitat. Most of these chameleons Avere taken on low bushes or 
brambles growing in rank grass. Under the circumstances it seems 
strange that grasshoppers do not figure more prominently as an 
article of diet, the caterpillars which they prefer are mostly Lycaenid 
species, the beetles do not include many large kinds but a great variety 
of small and often brilliantly colored species. It is probable that flies 
figure less than is really the case for being delicate they are usually 
masticated beyond recognition. 



338 bulletin: museum of comparative zoology 

Chamaeleon fulleborni Tornier 

Chamaeleon fiilleborni Tornier, 1900, Zool. Jahrb. Syst., 13, p. 614, fig. H: Slope 
of Ngosi or Poroto Mtns. etc., Tanganyika Territory. 
137 (M. C. Z. 31286-335) Nyamwanga, Poroto Mtns. 17. iii. 30. 

Distribution. The above series are topotypes for Xyamwanga is the 
last village on the way up to the Ngosi Volcano in the Poroto Moun- 
tains. The species is only known from the type series one of which 
was said to be from "Nonde Nike" which presumably means Ukonde- 
Unyika, i.e. the district inhabited by the Wakonde and Wanyika 
tribes who are settled round about and in the Poroto Range. The 
third cotype came from Kungura Mountain which is presumably in 
the same range or vicinity. 

Variation. The series consists of 72 males, 59 females and 6 young. 
The tails of 25 males range from .47 to .54 of the total length, with an 
average of .51 ; those of 20 females from .48 to .54 with an average of 
.51; those of the 6 young average .48. In all the young the tail is 
from 2 to 3 mm. shorter than the length from snout to vent. 

Measurements. The largest male measures 222 (105+117) mm.; 
the largest female 204 (99+105) mm. 

Breeding. In ten females examined the eggs were still spherical 
varying })etween 6 and 7 mm. in diameter. The number of eggs de- 
veloping in these ten chameleons was 11, 11, 12, 14, 15, 15, 16, 18 and 
24 which gives an average of 15. The six young, evidently from the 
last breeding season, range from 70 to 77 mm. in length. 

Diet. The stomachs of all ten specimens examined held nothing but 
beetle remains, chiefly those of a large species of rose beetle and a 
conspicuous ladybird (coccinellid) ; it is unusual for chameleons to re- 
strict themselves to one form of diet and it may well be that the time 
of my visit coincided with the swarming of beetles. 

Chamaeleon werneri averneri Tornier 

Chamaeleon werneri Tornier, 1899, Zool. Anz., 22, p. 258, fig. 1 (of a horned 9 )"• 
" Maschona-Gebiet, Deutsch-Ost-Af rika " (presumably the Wamashonde 
district, Tanganyika Territory); Nieden, 1913, Mitt. Zool. Mas. Berlin, 
5, p. 98: Rufigi, Uhehe, Ufipa and Usagara. 

2 (M. C. Z. 31336-7) Mufindi East, Uzungwe Mtns. 9. i. 30. 

8 (M. C. Z. 31338-43) Kigogo, Uzungwe Mtns. 13-30. i. 30. 

Distribution. When Goetze collected the type which Avas a one- 
horned female, the region had but recently been brought into subjec- 
tion to German rule and I am not certain whether the Uzungwe 



loveridge: African herpetology 339 

Mountains formed part of the Wamashonde district, or whether the 
type locaHty should be looked for in the mountainous country just 
south of the Uzungwe. Possibly the type female came from the 
Uzungwe for a few months later Tornier figured the three-horned male 
collected by Goetze in the Uzungwe range which is generally regarded 
as the type locality of the form. 

J'ariation. The tails of the 5 males range from .49 to .53 of the 
total length, with an average of .51, those of the 5 females are from 
.48 to .51 with an average of .49. 

Coloration in life, cf Mufindi. A very richly colored chameleon 
but the species changes color much more rapidly than most members 
of the genus. The horns are Indian-red; there are four dark saddle- 
like markings on the back, the intermediate areas being buff; the sides 
are rich green mottled with black and a little white, the interstitial 
skin is a deep crimson-lake which shows in streaks as the animal in- 
flates. The greens are reminiscent of shades of lichen but the effect 
of all the coloring is to give a velvety appearance to the chameleon. 

9 Mufindi. Above green bice, darker on the back, the interstitial 
gular skin Indian -red. When annoyed this reptile changes from dark 
olive to muddy-black while several darker streaks appear on the 
occipital flaps and three rather indistinct ^'ertical bars upon the sides 
of the body. 

M easurevients . The largest male measures 201 (103+98) mm., the 
largest female 203 (100+103) mm. 

Breeding. The ova in all five females is undeveloped. 

Diet. The distinguishable contents of the stomachs was as follows : 
(i) Beetles; caterpillar; grasshopper, (ii) beetles; four caterpillars, 
(iii) many beetles, chiefly weevils; seven caterpillars, mostly Lycaenid; 
a few flying ants; large flies; muscid fly; one froghopper, (iv) beetles; 
caterpillar; froghopper, (v) beetles, including golden beetle; flies, 
including a greenbottle; other insects, (vi) beetles; a wasp; a small 
polydesmid. 

Parasites. Nematodes {Strongyluris brevicaudata) invariably present 
in the intestine but not in the stomach. 

Enemies. The horn of one of these chameleons was recovered from 
the stomach of a Mountain Buzzard {Buteo oreophilus) shot at Kigogo. 

Chamaeleon werneri dabagae Loveridge 

Plate 3, fig. 5 

Chamaeleon werneri dabagae Loveridge, 1932, BuU. Mus. Comp. Zool., 72, p. 
379: Dabaga, Uzungwe Mountains, Tanganyika Territory. 



340 bulletin: museum of comparative zoology 

5 (M. C. Z. 31344-8) Dabaga, Uzungwe Mtns. 1. i. 30. 

Variation. The tails of the 3 males range from .50 to .53 of the total 
length; those of the 2 females are .51. 

Parasites. Nematodes {Sirongyluris brevicaudata) were present in 
the intestines. 

Chamaeleon jacksoni vauerescecae Tornier 

Chamaeleon jacksoni var. vauerescecae Tornier, 1903, Zool. Jahrb. Syst., p. 176: 
Nairobi, Kenya Colony. 

2 (M. C. Z. 31365-6) Nairobi, Kenya Colony. 5. vii. 30. 

Variation. The tails of these 2 females are .50 and .53 of the total 
length. 

Measurements. The larger measures 208 (104+104) mm. 

Chamaeleon incornutus Loveridge 

Plate 3, fig. 4 

Chamaeleon incornutus Loveridge, 1932, Bull. Mus. Comp. Zool., 72, p. 380: 
Madehani, Ukinga Mountains, southwestern Tanganyika Territory. 

12 (M. C. Z. 31350-5) Madehani, Ukinga Mtns. 14-28. ii. 30. 

4 (M. C. Z. 31356-9) Nyamwanga, Poroto Mtns. 17. iii. 30. 

5 (M. C. Z. 31360-4) Nkuka Forest, Rungwe Mtn. 19-30. iii. 30. 

The diagnosis of this species has been published alread}', a more 
detailed description follows. 

Description. Prefrontal region between the canthi rostrali flat and 
covered with large tile-like plates; canthi rostrali sharp, furnished 
with knob-like scales over the supraocular region to the occipital 
flaps; the latter also covered with smooth tile-like plates; around the 
upper part of the eyelid are long soft spines which are very conspicuous 
in life but apt to become flattened in preserved specimens; sides and 
upper surface of limbs covered with conical granular scales, enlarged 
ones are scattered among the smaller on the flanks but there are so 
many enlarged on the forearm as to present the appearance of armor- 
plating; the lower (inner) surface of the limbs is covered with fine 
granules. 

Variation. The tails of the 10 males range from .45 to .50 of the 
total length, with an average of .48; those of the 10 females are from 
.48 to .51 with an average of .48. 

Coloration in life. Paratype cf . Xkuka Forest. Above, crown of head 



loveridge: African herpetology 341 

light red as also a broad vertebral band to the end of the body; on 
either flank a narrow straight band extends from the occipital lobes 
to the end of the body; rest of body rich sap-green except for a narrow 
line from the throat to the end of the tail along the median line of the 
belly which is, like the inner sides of the limbs, a dirty white. 

Measurements. Type cf. Snout to anus 84 mm. Length of tail 77 
mm. 

The largest male measures 186 (93+93) mm.; the largest female 
188 (93+95) mm. 

Breeding. All the Madehani and Nyamwanga females hold eggs 6 
mm. in diameter and varying in number from 11 to 16 (3 only counted) ; 
two females from the Xkuka Forest would appear to have been taken 
during oviposition for each has only three eggs in the oviduct, these 
eggs measure 7 mm. A young chameleon taken in the Xkuka Forest 
measures 77 (41+36) mm. 

Diet. The distinguishable contents of ten stomachs examined was: 
(i) Beetles, including two large cetonids, a coccinellid and a tortoise 
beetle; also a caterpillar, (ii) a hairy and a smooth caterpillar, (iii) 
wings of dipteron; two hairy and one smooth caterpillar, (iv) cock- 
roach, (v) beetles; fly; attid spider; snail, (vi) beetle; caterpillar, (vii) 
beetle; homopteron; spider, (viii) beetle; homopteron. 

Parasites. Nematodes {Strongyluris brevicaudata) were collected 
from the intestines. 

Habitat. I caught the type as it was descending the lichen-covered 
trunk of a tree a hundred yards from the forest-edge. Recognizing 
it as a species new to the Territory it was shown to the natives and a 
special reward of ten cents (2 cents in U. S. currency) offered. Un- 
doubtedly it is a scarce forest-glade form as only four of over one hun- 
dred and fifty chameleons brought in at Nyamwanga were of this 
species. 

Chamaeleon laterispinis Loveridge 
Plate 3, fig. 3 

Chamaeleon laterispinis Loveridge, 1932, Bull. Mus. Comp. Zool., 72, p. 381: 
Kigogo, near Mufindi, Uzungwe Mtns., Tanganyika Territory. 
3 (M. C. Z. 31386-8) Kigogo, Uzungwe Mtns. 18-23. i. 30. 

The diagnosis of this species has been published already, a more 
detailed description follows. All three types are males. 

Description. Casque moderate, rather low, no rostral or other 
processes; canthi rostrali terminating in two scales; a rather indistinct 
parietal crest forking anteriorly; occipital flaps movable but somewhat 



342 bulletin: museum of comparative zoology 

rudimentary, covered by large flat as well as knob-like plates; a 
prominent thorn-like scale in the temporal region and others on the 
eyelid above the eye; a beard-like appearance is produced on the 
throat by reason of the many long scales anteriorly and on the sides. 
Vertebral line with a series of 17 large, soft spikes, each equidistant 
from its fellow, separated by an interspace on the base of the tail 
from another, but smaller and gradually dwindling, series on the tail; 
sides covered with large, flat, or rounded plates which are separated 
from one another by smaller granular scales; on either side of the body 
are two (or three) groups of soft, thorn-like spines, each group com- 
posed of 2 (occasionally 1 or 3) rather flattened scales about 2 mm. in 
length ; a further series of these thorn-like spines on either side of the 
tail and many others scattered over the limbs. Below, the hinder end 
of the throat, median line of the belly and underside of the tail are 
small granules. 

Coloration in life. Paratype cf . Above, pale green, excepting for the 
dorsal area and top of head which is a pinkish-brown with black saddle 
markings; the paired "thorns" pinkish-brown but changing to white 
while the black hour-glass -like markings mentioned in the type become 
deep green. The effect of this coloration is to produce a very lichen- 
like appearance. 

The type cf changed from the description given, as follows. The 
greenish-white became quite green, the black markings became 
speckled with brown and a good deal of reddish-brown appeared, 
chiefly along the vertebral line and on the tail. 

Diet. Numerous small beetles, heads of flies, a termite and three 
caterpillars were found in the stomach of one of these chameleons. 

Parasitrs. The intestine was full of nematodes {Strongyluris brcvi- 
caudata) but there were none visible in the stomach. 

Habitat. All three specimens were taken by Salimu or myself close 
to our camp above the forester's house where they are to be found on 
the small shrubs at the edges of the little patches of forest. As only 
three were secured during the three weeks spent at Kigogo it appears 
highly probable that the species is rare, at least in this locality. 

Brookesia temporalis (Matschie) 

Chaniaeleon (Brookesia) temporalis Matschie, 1892, Sitz. Ges. Naturf. freunde 
Berlin, p. 108: Derema, Usambara Mtns., Tanganyika Territory. 

Chaniaeleon temporalis Tornier, 1897, Kriechthiere Deutsch-Ost-Afrikas. p. 62, 
pi. ii. fig. 5; Werner, 1902, Zool. Jahrb. Syst., 15, p. 393; Werner, 1911, 
Das Tierrich. Chamaeleontidae, p. 23. 



LOVERIDGE: AFRICAN HERPETOLOGY 343 

1 (M. C. Z. 24385) Amani, Usambara Mtns. 24. xi. 26. 

Affinities. As already stated in the discussion on the generic status 
of Rhampholeon under the heading of Chamaeleonidae this interesting 
little species has long been lost sight of owing to Tornier's mistaken 
action in referring it to the genus Chamaclcon. 

In connection with the identification of the specimens of ijlaiijceps 
next following, it occurred to me to reexamine M. C. Z. 24385 which 
was obtained at Amani, only three miles distant from Derema, and 
which in 1928 had been referred to brevicaudatus with a long series of 
that species. I found that it agreed in every detail with Matschie's 
description of temporalis including the simple claws, and smooth scales 
on the soles of its feet. 

Brookesia platyceps (Giinther) 

Rhampholeon platyceps Giinther, 1892, Proc. Zool. Soc. London, p. 556, pi. 
xxxiv, fig. 1 : Shire Highlands, Nyasaland. 

2 (M. C. Z. 31367-8) Tandala, Ukinga Mtns. 11. ii. 30. 
1 (M. C. Z. 31369) Madehani, Ukinga Mtns. 13. ii. 30. 

3 (M. C. Z. 31370-2) Nyamwanga, Poroto Mtns. 17. iii. 30. 

4 (M. C. Z. 31373-6) Ngosi Volcano, Poroto Mtns. 19. iii. 30. 
21 (M. C. Z. 31377-83) Nkuka Forest, Rungwe Mtn. iv. 30. 

Distribution. B. platyceps is only known from the type from the 
Shire Highlands and a male from Nyasaland; the finding of the same 
species in the Ukinga and Poroto Mountains to the north and north- 
west of the lake is quite in accord with the occurrence of mammals, 
birds, and some reptiles such as Lygodactylus angularis in the same 
mountains. 

Native names. Wingiruli (Kikinga) ; haniula (Kisufi) ; katumbasagesi 
(Kinyakusa). 

Affinities. B. platyceps forms one of a group (which includes mar- 
shalli of Chirinda Forest, S. Rhodesia and boulengeri from northwest 
of Lake Tanganyika) characterized by a small scaly flexible, rostral 
process, bicuspid claws and tubercular soles but no digital spine. 

One of the series was submitted to Mr. H. W. Parker for favor of 
comparison with the type of Rhampholeon platyceps. He %\Tites: "As 
regards the Rhampholeon compared with the type of platyceps, your 
reptile has a more pronounced nasal appendage, larger supraciliary 
horns, rather flatter head and the upper head scales flatter. Another 
specimen, a male from Nyasaland, has a longer nasal appendage than 
yours, but the supraciliary horns are about the same size. Head scales 



344 bulletin: museum of comparative zoology 

are as in the type. In default of any other comparative material I 
should be inclined to regard them as conspecific or at most racially 
distinct." 

Variaiion. The tails of the 13 males are from 0.21 to 0.27 of the 
total length, with an average of 0.24; those of the 18 females 0.20 to 
0.24 with an average of 0.21. 

Coloration. A Tandala female was wood-brown while the Madehani 
female was sap-green w4th brownish areas, chiefly the limbs and head. 
Young from Nyamwanga were cream-colored or bamboo-white. 
Of those from Ngosi Volcano it was noted that all had the circular 
eyelids spotted with bright, yet pale, blue. 

Measurements. The largest male measures 53 (41 -(-12) mm.; the 
largest female 76 (60-|-16) mm., adults of the latter are always much 
larger than those of the former. The smallest specimen measured 42 
(31-hll)mm. 

Breeding. Ova small but developing at ^Vladehani on February 13th ; 
of the Ngosi series one held 12 eggs measuring 3 mm. in diameter, 
another 11 eggs of 4 mm., and a third 9 eggs of 5 mm. on March 19th. 
Of two females brought into camp at Rungwe on April 5th the smaller 
(55 mm. in total length) held numerous ova 2 mm. in diameter, the 
larger (70 mm.) held a single ver}- big egg measuring 12 x 7 mm. with 
what is apparently the commencement of an embryo in it. I con- 
cluded that she had probably been taken by a native when in the act 
of laying, but this is supposition. 

Diet. In the stomachs examined the following were found: — (1) 
Caterpillar and many termites, (2) beetle and grasshopper, (3) beetle, 
grasshopper, spider, (4) very small grasshoppers and spider, (5) cater- 
pillar, beetle, bug, (6) several spiders. 

Parasites. Some cestodes recovered from two chameleons taken on 
Ngosi Volcano have been identified by Baer as Neriiatotaenia jager- 
skidldi Janicki. Unfortunately in reporting upon them, that author 
(1933, Revue Suisse Zool., 40, p. 79) followed my field label in giving 
the host as Rhampholeon brevicaudatus (Matschie). Matters are 
further confused by his giving the locality as "Rhodesie meridionale," 
the material passed through other hands before reaching Dr. Baer 
and the label may have been copied erroneously. 



loveridge: African herpetology 345 

PT. II. Amphibia 
MATERIAL 

The period of collecting was from October 27, 1929 to July 9, 1930 
during which time 2,759 amphibians representing 60 species were pre- 
served. This total comprised 2 species of caecilians, 6 of toads and 
52 kinds of frogs ; of these 20 were new to the collection of the Museum 
of Comparative Zoology. 

This amphibian collection is somewhat disappointing, the subtropi- 
cal rain forest and rain-swept plateau of the southwestern highlands 
is much poorer in number of species than the more tropical mountains 
to the north. Still I was successful in getting topotypic series of nine 
species which were desired. Mention might be made also of the follow- 
ing rarities: Scolecomorphus kirkii, Rana floweri, Hyperolius rhodos- 
celis, Arthroleptis reichei, A. schubotzi and A. moorei. 



SUMMARY OF TAXOXOMIC ALTERATIONS 

The following species or races from this collection have been de- 
scribed briefly^ ; additional information regarding them will be found 
in the present paper; another is described beyond. 

Boulengerula changamwensis Changamwe, near Mombasa, Kenya. 

Bufo urunguensis Kitungulu, Urungu, Tanganyika. 

Bufo parkeri Mangasini, Usandawi, Tanganyika. 

Bufo taitamis uzunguensis Uzungwe, Ubena and Paroto Mtns. 

Rana mascareniensis uzungwensis Dabaga, Usungwe Mtns., Tanganyika. 

Arthroleptis rungwensis Ilolo, Rungwe district, Tanganyika. 

Arthroleptis ukingensis Madehani, Ukinga Mtns., Tanganyika. 
Probreviceps macrodactylus rungwensis Rungwe Mtn., Tanganyika. 

Hyperolius parkeri sp. nov. Dar es Salaam, Derema, Bagamoyo. 

In addition to the new species, the following are recorded from 
Tanganyika Territory for the first time : 

Scolecomorphus kirkii Boulenger. Type locality doubtful. ? Nyasaland. 

Rana floweri Boulenger of the Sudan and Mozambique. 

Rana ansorgii Boulenger of Angola. 

Phrynobatrachus perpalmatus Boulenger of the Belgian Congo. 

Hyperolius marginatus Peters of Mozambique, also recorded by Ahl in 1931. 

Loveridge, 1932, Bull. Mus. Comp. Zool., 72, pp. 375-387. 
Loveridge, 1932, Occ. Papers Boston Soc. Nat. Hist., 8, pp. 43-54. 



346 bulletin: museum of comparative zoology 

while the undermentioned are revived and should be added: 

Xenopus vidorianus Ahl as Xenopus laevis vidoriavus Ahl. 

Arthroleptis whytii Boulenger from the synonymy .>f A. stenodadylus Pfeffer. 

Hemisus guineensis Cope as Hemisus marmoratus guineensis Cope. 

The following are considered strict synonyms: 

*Rana barbouri Loveridge = Rana floweri Boulenger 

*Abrana colli Parker = Rana (Abrana) floweri Boulenger 

Rana Iheileri Mocquard = Rana oxyrhynchus Smith 

*Rana fiilleborni Nieden = Rana fasciala merumontana Lonnberg 

Phrynobalrachus p. werneri Ahl = Phrynobatrachus perpalmalus Boulenger 

Chiromanlis pygmaeus Ahl = Chiromanlis petersii pelersii Boulenger 

*Chiromanlis pidus Ahl = Chiromanlis petersii petersii Boulenger 

Chiromanlis rugosus Ahl = Chiromanlis petersii pelersii Boulenger 

Hylambales brevipahnalus Ahl = Leptopelis bocagii (Giinther) 

*Leplopelis barbouri Ahl = Leptopelis aubryi (A. Dumeril) 

*Leptopelis tangayius Ahl = Leptopelis uluguruensis Barbour & Lov- 
eridge 

*Leplopelis signifer Ahl = Leptopelis vermiculalus (Boulenger) 

Megalixalus dorsimaculalus Ahl = Megalixalus fornasinii (Bianconi) 

Hyperolius pygmaeus Ahl = Megalixalus fornasinii (Bianconi) 

1 Hyper olius inultifasciatus Ahl = Megalixalus brachynemis Boulenger 

IHyperolius acuticeps Ahl = Megalixalus brachynemis Boulenger 

Hyperolius ipianae Ahl = Megalixalus brachynemis Boulenger 

Hyperolius unicolor Ahl = Megalixalus brachynemis Boulenger 

IHyperolius asper Ahl = Hyperolius syinetricus Mocquard 

*Hyperolius ferniquei Mocquard = Hyperolius striolatus Peters 

Hyperolius coeruleopunctatus Ahl = Hyperolius striolatus Peters 

Hyperolius udjidjiensis Ahl (part) = Hyperolius striolatus Peters 

Hyperolius substriatus Ahl = Hyperolius puncticulalus (Pfeffer) 
Hyperolius callichromus Ahl (part) = Hyperolius puncticulalus (Pfeffer) 

^Hyperolius pidus Ahl (part) = Hyperolius marginatus Peters 

*Hyperolius ngoriensis Ahl = Hyperolius marginatus Peters 

* Hyperolius fuelleborni Ahl = Hyperolius mariae Barbour & Loveridge 

It is with real regret that the necessity is forced upon me of relegat- 
ing to the synonymy more than a score of species very recently de- 
scribed by Dr. Ernst Ahl. Either this author's concept of a species is 
at variance with that held by most herpetologists or it would appear 
that he has resorted to a "mass production" method of speciation by 
describing all material which was not readily identifiable. Such a 
method tends to throw herpetology into chaos and transfers to others 
the burden of discovering the true taxonomic status of the forms de- 

*Type or topotype examined. 



loveridge: African iierpetology 347 

scribed. As an example of what I mean by mass production I would 
cite a recent paper by Dr. Ahl (1931, Mitt. Zool. INIus. Berlin, 17, pp. 
1-132) in which he describes as new ninety-eight "species" of the genus 
Hyperolius or considerably more than all authors combined have de- 
scribed during all time. In his eagerness for description he does not 
hesitate to describe frogs "Ohne genauen Fundort" or with "Africa" 
only designated as the type locality. Little effort appears to have been 
expended on the correct spellings of place names and this has resulted 
in names like Hyperolius ngoriensis given to the juvenile frog of his 
H. pidus both being attributed to "Krater des Ngori-See's," really 
the Crater Lake of Xgosi Volcano. Designations which have been 
obsolete for many years, have been employed for larger areas, thus 
Abyssinia is used instead of Ethiopia; British-Ostafrika for Kenya 
Colony; Deutsch-Ostafrika for Tanganyika Territory as well as for 
Belgian Ruanda-Urundi; Portuguese Ost-Afrika for Mozambique; 
these are just a few of the changes which have been ignored in the 
compilation of Amphibia Anura iii in Das Tierrich, 1931, 55. 

In addition to the species synonymized above, a number of corrected 
records will be found in the bibliography of certain species. 

Spawn or tadpoles were collected of Xenopus laevis victoriami.s , X. 
muclleri, Rana fuscigula angolensis, Rana galamensis, Arthrolcptis 
parvulus, Chiromantis petersii petersii, Hyperolius viarginatus, H. 
parkeri, Hemisus m. marmoratum and breeding conditions of other 
species are noted. 

ACKNOWLEDGEMENTS 

Once again I wish to thank Mr. H. W. Parker for comparing various 
specimens with the types in the British Museum collection, his com- 
ments are included in the text. Also I am indebted to Mons. Angel 
for affording me the opportunity to examine certain old types of the 
genus Hyperolius of whose identity I was in doubt. The resultant ob- 
servations are published in this paper though the species may not 
have been collected during the course of the expedition. 

List of Species Collected* 

Order Apoda page 

Family CAECILIIDAE 350 

Scolecomorphus kirkii Boulenger 350 

Boulengerula changamicensis Loveridge 351 

*An asterisk opposite a species indicates that examples are available for exchange. Species 
In parentheses are discussed though not collected. 



348 bulletin: museum of compaeative zoology 

Order Salientia page 

Family PIPIDAE 351 

*Xenopus laevis vidorianus Ahl 351 

*Xeno2)us poweri Hewitt 352 

*Xenopus muelleri Peters 353 

Family BUFONIDAE 354 

*Bufo regularis regularis Reuss . . 354 

Bufo carens Smith 355 

Bufo unmguensis Loveridge 356 

*Bufo parkeri Loveridge 356 

Bufo taitanus uzimguensis Loveridge 357 

*Nectophrynoides vivipara (Tornier) 357 

Family RANIDAE 359 

*Rana adspersa (Dumeril & Bibron) 359 

*Rana delalandii (Dumeril & Bibron) 359 

*Rana occipitalis Giinther 361 

Rana fuscigula chapini Noble 361 

Rana fuscigula angolcnsis Bocage 362 

Rana galamensis Dumeril & Bibron 365 

*Rana floweri Boulenger 367 

*Rana oxyrhynchus Smith 368 

*Rana mascarcniensis mascarenicnsis Dumeril & Bibron .... 369 

*Rana mascarcniensis uzungicensis Lo^'eridge 370 

Rana viascareniensis venusta Werner 370 

*Rana ansorgii Boulenger 371 

*Rana fasciata mcrumontana Lonnberg 372 

*P}iryuobafrachus natalensis (Smith) 373 

*Phrynohairachus acridoidcs (Cope) 374 

Fhrynohatrachus pcrpalmatus Boulenger 375 

*Arthroleptis stenodactylus stcnodactylm Pfeffer 376 

*Arthroleptis ichytii Boulenger 378 

*Arthrolcptis adolfi-fricdcrici Nieden 379 

*Arfhrolrpfis rcichci Nieden 379 

* Art h role ptis schubotzi Nieden 380 

Irthrolcptis xenodaciylus Boulenger 382 

'^Arthroleptis moorii Boulenger 383 

Artkroleptis minnius Boulenger 384 

Arthrolepfis ukingensis Lo\'eridge 385 

*Aii asterisk opposite a species indicates that examples are available for exchange. Speciej 
in parentheses are discussed though not collected. 



* 
* 
* 



loveridge: African herpetology 349 

PAGE 

Arthrolcptis rimgwetisis Loveridge 386 

*Arfhrnlrptis parvulus Boulenger 386 

*IIrniisus niarmoratum marmoratmn (Peters) 388 

llcmisus marmoratum guincensis Cope 389 

Family POLYPEDATIDAE 390 

^Cldromantis piicrsii pctersii Boulenger 390 

Lcptopelis hocagii (Giinther) 393 

Leptopelis joknstoni (Boulenger) 394 

Lcptopelis rermiculatus (Boulenger) 395 

(Lcptopelis auhri/i (A. Dumeril)) 396 

(Lcptopelis uhiguruensis Barbour & Loveridge) 397 

*Mcgalixalus fornasinii (Bianconi) 397 

*MegalixaIus brachynemis Boulenger 399 

Hyperolius ? sansibaricus (Pfeffer) 399 

(Hyperolius viridifiavus (Dumeril & Bibron)) 400 

(Hyperolius symetricus (^Vlocquard)) 400 

(Hyperolius striolatus Peters) 401 

*Hypcrolius calUchromus Ahl 403 

*Hyperolius rhodoscelis (Boulenger) 404 

Hypjcrolius picturatus Peters 405 

^Hyperolius puncticidatus (Pfeffer) 406 

^Hyperolius marginatus Peters 406 

"Hyperolius mariac Barbour & Loveridge 409 

Hyperolius ? platyrhinus (Procter) 409 

Hyperolius gramdatus (Boulenger) 410 

'^Hyperolius parkcri sp. nov 410 

^Kassina senegalensis (Dumeril & Bibron) 412 

Family BREVICIPITIDAE 413 

*Brcricejjs mossambicus Peters 413 

Probreviceps macrodactylus rungwensis Loveridge 414 

(Hoplophryne uluguruensis Barbour & Loveridge) 414 

*P}irynomeriis bifasciatus (Smith) 415 

*An asterisk opposite a species indicates that examples are available for exchange. Species 
in parentheses are discussed though not collected. 



* i 



*J 



350 bulletin: museum of comparative zoology 

Systematic List of Species Collected 

CAECILIIDAE 

ScoLECOMORPHUS KiRKii Boulenger 

Scolecomorphus kirkii Boulenger, 1883, Ann. Mag. Nat. Hist. (5), 11, p. 48: 
"Probably vicinity of Lake Tanganyika." 

1 (M. C. Z. 16305) Mufindi-Njombe Road, Ubena. 6. ii. 30. 

Distribution. Though the type locality of the species was uncertain 
Boulenger subsequently recorded several examples from Nyasaland; 
I anticipated finding it in southwestern Tanganyika Territory {vide. 
Loveridge, 1930, Proc. Zool. Soc. London, p. 10, footnote and key to 
species). 

Native navies. Timagivini (Kibena); melaundetzi (Kikinga). 

Variation. This specimen has 152 annular rings (or 157 if some 
incomplete and indistinct ones are reckoned) as had the type accord- 
ing to Boulenger (though in its present somewhat softened condition 
I counted 149). There are now eight examples in the British Museum 
from Zomba and the Shire Highlands and these range from 134-149, 
only one Zomba specimen is as low as 134, the series otherwise being 
142-149, the range may be regarded as 134-152. 

Coloration in life. Above olive; below pinkish-white. After a year 
in formalin and alcohol it is blue grey above and flesh-colored below, 
i.e. exactly like *S. uluguruensis preserved by the same methods. 
Boulenger described the alcoholic type as, "Dark olive above, brown- 
ish olive beneath." 

Measurements. When freshly killed the length was 342 mm., the 
diameter 10 mm., the latter being contained in the former 34.2 times; 
as now preserved the length is 326 mm. and the diameter 9 mm., the 
latter being contained in the former 35 times. The type was 38.5 
times according to Boulenger, 35 times at the present day; the range 
of the British Museum series is from 30 to 41 times with a 3'oung 182 
mm. specimen of 51.1 times. The range for the species is therefore 
30 to 51.5 times. 

Habitat. When proceeding from Mufindi to Njombe our lorry be- 
came stuck several times in the black cotton soil. When about sixty 
miles out from Mufindi this occurred again; in digging out the back 
wheels one of my boys unearthed this fine caecilian. The surrounding 
country was more or less open grassland plateau with scattered orchard 
forest of small trees. 



loveridge: African herpetology 351 

BOULENGERULA CHANGAMWENSIS Loveridge 

Boulengerula changamwensis Loveridge, 1932, Bull. Mus. Comp. Zool., 72, p. 
381 : Changamwe, near Mombasa, Kenya Colony. 

4 (M. C. Z. 16301-4) Changamwe, K. C. 31. x. 29. 

Habitat. The envii-onment in which these caeciHans were found is 
totally different from that of B. uluguniensis, the altitude of Chan- 
gamwe station is only 191 feet. I was engaged in turning over a heap 
of weeds beneath a mango tree when I caught sight of the first caecilian. 
The soil below was black earth with a liberal admixture of sand; round 
about the soil was composed almost entirely of sand. Earthworms 
were abundant enough under the mango trees but an hour of strenuous 
digging in the vicinity only resulted in securing tliree more caecilians. 

PIPIDAE 

Xenopus laevis victorianus Ahl 

Xenopus victorianus Ahl, 1924, Zool. Anz. Leipzig, 60, p. 270: Bussisi, i.e. 

Busisi, Lake Victoria, Tanganyika Territory. 
Xenopus laevis Loveridge (part), 1925, Proc. Zool. Soc. London, p. 766. 

1 (M. C. Z. 16329) Mwanza, Lake Victoria. 6. vi. 30. 
Tadpole and 9 (M. C. Z. 16320-8) Kampala, Uganda, vi. 30. 

The Mwanza frog, collected just across the gulf from Busisi, may 
be considered almost topotypic of Ahl's victorianus, a species which in 
1925 {loc. cit.) I assumed to be a strict synonym of laevis. The reasons 
given for the opinion then expressed are still valid, that is to say, the 
structural characters on which its author differentiated his holotype 
from laevis are all variable and the variations common to laevis-. 
With more material, however, I now find that on size, average color 
of adults, and shape of claws, victorianus may be recognized as a race 
of laevis so I retract the view held in 1925. 

The relations of the East African members of the short -tentacled 
laevis group may be summarized as follows : 
Habit p\Tiform, the greatest width of the head 

being included from one and a third to once and 

two-thirds times in the greatest width of the 

body. 

Size large (103 mm. maximum for 33 speci- 
mens); black claws flattened when viewed 



352 bulletin: museum of comparative zoology 

from above; belly usually immaculate, 
rarely flecked or vermiculated with grey or 
brown A'. /. laevis 

Size moderate (65 mm. maximum for 14 speci- 
mens); black claws narrow, relatively slen- 
der when viewed from above; bellv usually 
flecked or spotted with grey or black, im- 
maculate in young A'. /. mctorianus 

Size moderate (65 mm. maximum for 10 
specimens); black claws narrow, rounded 
when viewed from above; belly always 
heavily vermiculated or blotched with 

black A', poweri 

Habit slender, sides parallel, the greatest width 
of the head being equal to, or only a trifle larger, 
than the greatest width of the body. 

Size small (53 mm. maximum for 173 speci- 
mens) ; black claws narrow, relatively 
slender when viewed from above; belly 
usually, thighs below almost always (2% 
are not), flecked and spotted with black. 
Breeds at 35 mm A'. /. hunyoniensis 

Xenopus poweri Hewitt 

Xenopus poweri Hewitt, 1927, Records Albany Mus., 3, p. 413; pi. xxiv, fig. 3: 
Victoria Falls, Northern Rhodesia. 

10 (M. C. Z. 16312-9) Tukuyu, Rungwe. 13. iii. 30. 

Affinities. The above series were submitted to ]Mr. J. Hewitt for 
favor of comparison with the type and only known specimen of 
poweri with which all agree in the heavy vermiculation of the lower 
surface. The two largest frogs measure 65 mm., the same as the type 
of poweri. 

Mr. Hewitt writes: "Yours is certainly near to my poweri: the 
latter, known only from the type specimen, unfortunately, has a trifle 
longer eye tubercle, otherwise I would say is the same species as yours. 
I reject them as laevis. If you compare their claws with those of 
laevis you will have no difficulty in distinguishing them. The claws 
of your specimens when unworn are long and slender, something like 
those of muclleri. Our specimens have flatter claws. Our laevis also 
grows bigger, never has well defined spots on the belly and I fancy the 



loveridge: African herpetology 353 

mettitursiil tubercle is stronger. Thus poirfri cannot go in lacins nor in 
mucUcri but I have sometimes considered if it can be a hybrid: your 
series has restored my confidence in it as a distinct entity. I have not 
investigated skeletal characters in your material but you will notice 
from my account of poivrri that the foot seems to be promising as a 
means of discrimination." Later he wrote agreeing that "the metatar- 
sal tubercle prominence is variable in appearance according to the 
preservation." 

I take this opportunity of expressing my thanks to Mr. Hewitt for 
these very helpful observations. Finding considerable variability in 
eye tentacle length in other species, however, I do not feel justified in 
describing this Tukuyu form as distinct from poweri on such slender 
grounds. Moreover Tukuyu is less than eight hundred miles from the 
Victoria Falls and they are connected except for a hundred miles by 
the Luangwa and Zambesi, the former Ijeing a tributary of the latter. 

After checking with the material in the Museum of Comparative 
Zoology, I have embodied Mr. Hewitt's views on the distinguishing 
characters of these frogs as opposed to lacvis in the key on the pre- 
ceding page. It might be added that the measurements given for 
A'. /. laevis are based on Kenya specimens for undoubtedly true lacvis 
occurs on the upland plateau of that colony while A'. /. ridoriamis not 
only occurs round the shores of Lake Victoria but also adjacent to the 
rain-forest outlyers of the Usambara and Uluguru Mountains. 

Xenopus muelleri (Peters) 

Dadylethra muelleri Peters, 1844, Monatsb. Akad. Wiss. Berlin, p. 37: Mozam- 
bique. 

6 (M. C. Z. 16306-8) Handa, Usandawi. 10. xii. 29. 
1 (M. C. Z. 16309) Sumbwa, Lake Tanganyika. 20. v. 30. 
Tadpoles (M. C. Z. 16310) Albertville, Lake Tanganyika. 21. v. 30. 
1 (M. C. Z. 16311) Shinyanga, Usukuma. 4. vi. 30. 

Coloration. All are spotted on the ventral surface. 

Breeding. The identification of the big series of tadpoles can be 
regarded as tentative until adult frogs are collected at Albertville. 
These curious tadpoles were swimming about in a swamp close to the 
lake shore. Below them and quite motionless with claws extended lay 
a Bellostoma apparently awaiting the near approach of an unwary 
pollywog. Young numerous at Handa. 

Habitat. The Handa frogs were taken from waterholes in a valley 
in very arid country. Only one adult was found, scores of young were 



354 



bulletin: museum of comparative zoology 



netted and released. The Sumbwa specimen was in a small pool close 
to the lake. The Shinyanga frog in a pool in the ri^'er bed, the whole 
region being very dry at the time of my visit. 



BUFOXIDAE 
BuFO regularis regularis Reuss 
Bufo regularis Reuss, 1834, Mus. Senckenberg, 1, p. 60: Egypt. 
1 (M. C. Z. 16.351) Miritini, K. C. 30. x. 29. 

1 (M. C. Z. 17079) Bagamoyo, T. T. 11. xi. 29. 

2 (M. C. Z. 16352) Mpwapwa, Ugogo. 23. xi. 29. 

4 (M. C. Z. 16353-4) Mangasini, Usandawi. 14. xii. 29. 

1 (M. C. Z. 16355) Dabaga, Uzungwe Mtns. 1. i. 30. 

1 (M. C. Z. 16356) Mwaya, Lake Nyasa. 1. iii. 30. 

4 (M. C. Z. 16357-8) Ilolo, Rungwe. 15. iii. 30. 

1 (M. C. Z. 16359) Nkuka Forest, Rungwe Mtn. iii. 30. 

1 (M. C. Z. 16360) Igale, Poroto Mtns. 30. iv. 30. 

1 (M. C. Z. 16361) Nr. Ikombo, N. Rhodesia. 6. v. 30. 

6 (M. C. Z. 16362) Nyamkolo, Lake Tanganyika. 9. v. 30. 

1 (M. C. Z. 16363) Kitungulu, Urungu. 15. v. 30. 

2 (M. C. Z. 16364-5) Albertville, Lake Tanganyika. 21. v. 30. 

1 (M. C. Z. 16366) Ujiji, Lake Tanganyika. 28. v. 30. 

15 (M. C. Z. 16367-71) Entebbe, Lake Victoria. 27. vi. 30. 

2 (M. C. Z. 16372-3) Kampala, Uganda, vi. 30. 
1 (M. C. Z. 16374) Jinja, Uganda. 30. vi. 30. 

1 (M. C. Z. 16375) Mabira Forest, Uganda. 1. vii. 30. 

Distribution. Also seen on the mainland opposite Kilindini harbour, 
and at Tanga, Mwandemeres and Tukuyu at which places toads were 
caught, examined and released. Noted as abundant in most of the 
localities except in the Uzungwe, Rungwe and Poroto Mountains. 

Native name. Ikiyula (Kinyakusa). 

J'ariation. There is the usual wide range of variability, the Igale 
toad was very warty and presented a strange appearance, the Kitun- 
gulu specimen was exceptionally spinose and so unusual in its colora- 
tion that I noted it in full as follows: 

Coloration in. life. Kitungulu. Above, a triangular area from the 
snout to between the eyes is pale buff and is followed by another area, 
bounded by the parotids, which is brick-red, upon its anterior part 
two irregular, sepia-hued blotches, two larger ones on its posterior 
portion, these are followed by a circular area of pale buff which also 
exhibits irregular markings in sepia on its posterior extremity, this 
area is surrounded by a darker red one extending nearly to the vent; 



loveridge: African herpetology 355 

sides of head and flanks pale buff vermieulated with black ; limbs pale 
pink barred with black. 

So striking and unusual was the coloring of this specimen that I 
could hardly bring myself to believe that it was only a Common 
Square-marked Toad. Moreover it was taken in a situation where one 
would have expected a dark and not a gaudy creature. 

Measurements. The largest specimen is a female from Mwaya 
measuring 94 mm., both a Mangasini and Dabaga toad are 91 mm. 

Breeding. At Bagamoyo, on November 11th, young ones were 
plentiful near waterholes close to the seashore, a toad taken at Mp- 
wapwa on the 23rd was only 11 mm. in length. 

At Mangasini, on December 14th, Square-marked Toads were 
calling "core-core" vociferously from the swamped flats. A frenzy of 
pairing was in progress, often several males were seen struggling for 
possession of one female; one male had clasped a female Rana adspersa. 
Both sexes of the toads were bright pinkish red on the hinder sides of the 
thighs, sometimes this color spread along the sides as far as the axilla. 

An unpublished note made at Kilindini on May 3, 1926, states that 
two pairs were taken in embrace and released. 

At Albert ville, on May 21, ditches outside the town were literally 
swarming with little toads recently emerged from the tadpole stage. 

Enemies. At Ilolo a young one was recovered from the stomach of a 
Rhombic Xight Adder {Causus rhoviheahis) and at Ujiji from a White- 
lipped Snake {Croiaphopeltis h. hotamhoeia). 

Habitat. Apart from those taken at waterholes, the following notes 
were made of occurrence of toads in other places. At Kilindini many 
were in holes a hundred yards from the shore, these were apparently 
crab holes from which the toads popped out their heads on hearing a 
footstep. At Tanga under mcuti or palm-frond thatching lying on the 
ground. At Mpwapwa in holes among the rotting roots of a tree 
stump. To my surprise I found several in open glades in the rain 
forest on Rungwe, some were in saw-pits in recent clearings, all were 
within half-a-mile of the forest edge and none in really dense forest. 
At Entebbe ten young ones w^ere dug out of a termite hill. 

BuFO carens Smith 

Bufo carens A. Smith, 1849, lllus. Zool. S. Africa, 3, pi. Ixviii, fig. 1: Interior 
of Southern Africa. 

1 (M. C. Z. 16387) Senjeri Pass, T. T. 5. v. 30. 
1 (M. C. Z. 16388) Near Ikombo, N: R. 6. v. 30. 

Habitat. Both were taken in roadside ditches. 



356 bulletin: museum of comparative zoology 

BuFO URUNGUENSis Loveridgc 

Bufo urunguensis Loveridge, 1932, Bull. Mus. Comp. Zool., 72, p. 383: Kit- 
ungulu, Urungu, Tanganyika Territory. 

5 (M. C. Z. 16376-9) Kitungulu, Urungu. 14. v. 30. 

Coloration in alcohol. Above, uniform grey broAvn. Below, white, 
spotted or vermiculated with dark brown in the pectoral region. 

Diet. Termites present in two of the paratypes examined. 

Habitat. The first example of this interesting little toad was caught 
by me when I was on the march to Kitungulu; there was just sufficient 
light for me to see that it was a species new to me; I carefully A\Tapped 
it in my handkerchief but on reaching camp after an hour's stumbling 
among stones in the darkness, found that it had escaped. Next day 
we searched for more and were successful in securing five on the 
swampy floor of a patch of primary forest beside the river. 

BuFO PARKERi Loveridgc 

Bufo parkeri Loveridge, 1932, Bull. Mus. Comp. Zool., 72, p. 382: Mangasini, 
Usandawi, Tanganyika Territory. 

30 (M. C. Z. 16330-50) Mangasini, Usandawi. 14. xii. 29. 

The diagnosis of this species has appeared already, the complete 
description follows : 

Description. Crown without bony ridge; snout short, rounded, with 
very distinct canthus; interorbital space, concave, equal in width to 
an upper eyelid; tympanum only fairly distinct, longer than broad, its 
breadth equal to half the diameter of the eye. Fingers rather pointed, 
first considerably shorter than the second; toes scarcely webbed at 
base, with both simple and paired subarticular tubercles and laterally 
with minute spines; two moderate metatarsal tubercles; no tarsal 
fold; the tibio-tarsal joint reaches midbody; the tarso-metatarsal 
tubercle reaches the tympanum. Upper parts with round warts of 
unequal size, a series of enlarged ones along the lateral line; parotids 
feebly prominent with a tendency to break up into warts; warts on 
limbs, hands and feet often terminating in minute spines. Male with 
a large A'ocal sac. 

Breeding. These toads were undouI:)tedly assembling for breeding 
in response to the breaking of the rains. 

Habitat. The first heavy rains fell at 5 p.m. on the 12th and lasted 
till noon on the 13th. On going down to the semi-flooded flats when 



loveridge: African herpetology 357 

it stopped I saw a great many small toads which at first I supposed 
to be young B. r. rcguJaris; as I was returning it struck me as curious 
that there were no intermediates in size between them and the adult 
regularis which were pairing in every pool; picking one up I observed its 
chrome colored throat and so arranged for Salimu to return the 
following day and secure a series. 



BuFO TAITANUS TJZUNGUENSis Loveridge 

Bufo taitanus uzunguensis Loveridge, 1932, Occ. Papers Boston Soc. Nat. Hist., 
8, p. 44: Kigogo, Uzungwe Mountains, Tanganyika Territory. 

2 (M. C. Z. 16380-1) Dabaga, Uzungwe Mtns. 1. i. 30. 
2 (M. C. Z. 16382-3) Kigogo, Uzungwe Mtns. 23. i. 30. 
1 (M. C. Z. 16384) Njombe, Ubena Mtns. 7. ii. 30. 
1 (M. C. Z. 16385) Lukungu, Ubena Mtns. 8. ii. 30. 
1 (M. C. Z. 16386) Nyamwanga, Poroto Mtns. 17. iii. 30. 

Distribution. Nieden, in 1913, recorded taitanus from near Iringa 
and from Rungwe Volcano so that I was on the lookout for this small 
earless toad. It must, however, be scarce, for only seven examples were 
encountered, most of these were taken by myself as the natives prob- 
ably confused them with the young of Bufo r. regularis which they 
closely resemble. 

Native names. TofuJa (Kihehe); Ikii/ula (Kinyakusa). 

Coloration in life. Kigogo. Above, brown, with a light yellow, hair- 
like line from snout to anus; a series of symmetrical black markings 
(very similar to those of B. r. regularis) on either side of this line; 
parotids chestnut-brown as also some of the lateral warts, otherwise 
sides grey almost obscured by large black patches ; limbs grey-brown 
above barred with black; a V-shaped cream-colored spot above the 
anus. Below, white, cream-colored, marbled with brown and black. 

Measurements. Length from snout to anus 21 to 29 mm., average 
25 mm. 

Habitat. Dabaga specimens were taken in swampy marshland beside 
a brook in the valley bottom. The Nyamwanga toad was hopping 
along on a sodden path which led through long grass up to the village. 



Nectophrynoides vivipara (Tornier) 

Pseudophryne vivipara Tornier, 1905, Sitzber. Akad. Wiss. Berlin, 2, p. 855: 
Rungwe and Ukinga Mountains, Tanganyika Territory. 



358 bulletin: museum of comparative zoology 

Nedophrynoides vimpara Barbour & Loveridge, 1928, Mem. Mus. Comp. Zool., 
50, p. 191: Uluguru Mountains, Tanganj-ika Territory. 

1 (M. C. Z. 16389) Kigogo, Uzungwe Mtns. 30. i. 30. 

2 (M. C. Z. 16390-1) Nyamwanga, Poroto Mtns. 17. iii. 30. 

56 (M. C. Z. 16392-400) Ngosi Volcano, Poroto Mtns. 18-20. iii. 30. 
28 (M. C. Z. 16401-9) Nkuka Forest, Rungwe Mtn. iii. 30. 

Distribution. The last listed specimens are topotypes. The Vivi- 
parous Toad is not nearly so abundant in any of these localities as it 
is in the Uluguru Mountains, it would appear to be even rarer in the 
Ukinga range than in the Uzungwe. 

Coloration in life. Young toads from Xgosi were flecked with silvery- 
white below and had the dorsal area edged with a silvery -white line 
which gave them the appearance of young Arthroleptis. 

Measurements. In the southwestern highlands the species does not 
appear to obtain to the large dimensions of specimens from the more 
tropical Uluguru. The largest from Xgosi Volcano is 48 mm., and the 
largest from the Nkuka Forest only 45 mm., not more than ten in the 
whole series are over 40 mm. 

Breeding. Most of the Ngosi series consists of small toads of 11 mm: 
in length, the smallest from Rungwe is 12 mm. At the same time three 
adult females from both Ngosi and Rungwe held large eggs as was the 
case with the Nyamwanga female but the ova were small in the Kigogo 
toad. 

Diet. Chiefly beetles, also cricket, caterpillar, spider and a wood- 
louse. 

Parasites. In Rungwe these toads exhibit a heavy infestation of 
larval mites which show as small red specks on hands and feet. 

Defence. When killed in chloroform the large glands on the back 
and limbs exude a considerable quantity of poison which is as fluid 
and white as cow's milk. 

Habitat. The Kigogo toad was taken under a rotting log at the edge 
of rain forest. At Ngosi one adult was taken four feet from the ground 
where it was climbing through undergrowth, another in grass a foot 
from the ground, se^'eral half-grown in bamboo and others up in the 
wild bananas, the majority, however, were among the dead leaves 
carpeting the forest floor. On Rungwe it was a rare occurrence to 
meet with a viviparous Toad below the bamboo belt but they were 
quite common on the path where it passes through the bamboos near 
the summit of the mountain. 



loveridge: African herpetology 359 

RANIDAE 
Rana adspersa (Dumeril & Bibron) 

Pyxicephalus adspersus Dumeril & Bibron, 1841, Erpet. Gen., 8, p. 444: South 
Africa. 

4 (M. C. Z. 16473-6) Bagamoyo. 11. xi. 29. 
12 (M. C. Z. 16477-86) Mangasini, Usandawi. 12-16. xii. 29. 

Distribution. Recorded by Nieden from Dar es Salaam, Kilimatinde 
and Unyika. 

Coloration, in life. Of two frogs, taken from the same swamp at 
Bagamoyo, the throat of one was handsomely marbled with olive but 
showed no yellow, in the other the throat, chest, belly and sides were 
bright yellow. 

Measurements. The Bagamoyo series range from 81 to 92 mm., 
those from Mangasini 94 to 138 mm., average of the whole series 110 
mm. 

Breeding. Dr. Rudolph Stohler has examined one of the Bagamoyo 
frogs and confirmed my opinion, based on two others, that they are 
immature. As they are not breeding one wonders why some of the 
series should have such bright coloring and others none. The deep 
note of adspersa was occasionally heard. The tadpoles found in the 
swamp are assumed to be those of galamensis. 

At Mangasini a female adspersa was found in the embrace of a male 
Bufo r. regularis. 

Diet. At Bagamoyo the stomach contents consisted of: (1) A cock- 
chafer and a tremendous number of tadpoles, (2) fewer tadpoles and 
three species of cockchafers. At INIangasini, (3) a lizard (Latastia 
johnstonii) and quantities of winged termites, (4) a frog, apparently a 
young R. adspersa, grasshopper, beetle and polydesmid, (5-7) termites 
and Megaponera. 

Parasites. An immature female ascarid was present in one Baga- 
moyo frog. 

Enemies. The remains of a large frog were recovered from the 
stomach of a cobra (Naja nigricollis). Apparently tadpoles and young 
are preyed upon by the adults as related above. 

Rana delalandii (Dumeril & Bibron) 

Pyxicephalus delalandii Dumeril & Bibron, 1841, Erpet. Gen., 8, p. 445, pi. 
Ixxxvii, figs. 1, la, & lb: South Africa. 



360 bulletin: museum of coSiparative zoology 

36 (M. C. Z. 16487-500) Masiliwa, Turu. 9. xii. 29. 
4 (M. C. Z. 16501-3) Mangasini, Usandawi. 14. xii. 29. 
1 (M. C. Z. 16504) Kikuyu, Ugogo. 21. xii. 29. 
1 (M. C. Z. 16505) Senjeri Pass, 5. v. 30. 

Variation. I find that the key character which I employed in 1930 
to distinguish this species from its allies does not hold for many of the 
specimens listed above which are bloated with termites; it would be 
better to say "the metatarsal tubercle of the adpressed hind limb 
reaches the eye" rather than that the tibio-tarsal articulation does so 
for the latter sometimes falls short. 

Coloration in life. In life the sexes may be very differently colored 
dorsally as at Masiliwa where females were red and males mottled and 
darker. At Mangasini a 37 mm. male was taken embracing a 45 mm. 
female and the upper surface of l)oth was essentially similar, viz. 
Above, greenish-white mottled and barred with dark green, such 
mottlings edged with black; markings, particularly on the head, like 
those of B. r. regularis; on back numerous round, chocolate-colored, 
black-edged spots. Beneath, however, the male's throat Avas greenish- 
black, that of the female like the rest of the undersurface — satiny- 
white with a slight greenish shading posteriorly. In alcohol the throats 
of males are black, those of females and immature frogs, white. 

Measurements. Sixteen black-throated males range from 28 to 44 
mm., average 34 mm.; omitting a few small specimens the twelve 
female adults range from 40 to 50 mm., with an average of 43 mm. 

Breeding. Mating at Mangasini on December 14th at which time a 
female was bloated with eggs. 

Diet. Those examined were gorged with flying termites, a few 
beetles were also found. 

Habitat. As a tent site was being cleared at Masiliwa at 3 p.m. I 
captured a single frog; about sunset several more were found hopping 
about capturing termites which were flighting after a heavy shower 
which fell between 4 and 5 p.m. It began to rain again and so I went 
out from 8 to 9 p.m. to some recently hoed-over ground and captured 
a great many more. In all eighteen of either sex. They were associated 
with other burrowing types in the same field, viz. Hemisus m. guineense 
and Breviceps mossamhicus. 

The Kikuyu frog was taken hopping about at night and the Senjeri 
Pass specimen at 11 p.m. in a roadside ditch; these frogs are essentially 
nocturnal and effectivelv conceal themseh'cs during the day. 



loveridge: African herpetology 361 



Rana occipitalis Giinther 

Rana occipitolis Giinther, 1858, Cat. Batr. Sal. Brit. Mus., p. 130, pi. xi: 
"West Africa," "Africa," Gambia. 

1 (M. C. Z. 16506) Nyamkolo, Lake Tanganyika. 9. v. 30. 

1 (M. C. Z. 16507) Kipili, Lake Tanganyika. 19. v. 30. 

7 (M. C. Z. 16508-14) Ujiji, Lake Tanganyika. 28. v. 30. 
3 (M. C. Z. 16515-7) Ukerewe Id., Lake Victoria. 10. vi. 30. 

2 (^L C. Z. 16518-9) Kampala, Uganda, vi. 30. 

Disirihution. Also seen at Shinyanga. Nieden remarks that this 
species is rare in German East Africa; it would seem tliat the eastern 
limit of its range is about a hundred miles east of Lake Victoria. 

Native navies. Chula (Kifipa) ; luncla (Kimanyema). 

Measurements. The Ujiji and whole series ranges from 25 mm. to 
1,300 mm., the latter, a female, being the only fully adult frog. 

Habitat. The big Ujiji frog was taken from a cement-lined pit 
twenty feet in depth as described under PcJuslos sinuatus in the report 
on the reptiles. The young one from Kipili was captured in a lagoon 
close to the lake shore. 



Rana fuscigula chapini Noble 

Rana nutii {nee. Boulenger) Andersson, 1911, Svenska Vetensk.-Akad. Handl., 
47, No. 6, p. 26: localities in Kenya Colony. Procter, 1920, Proc. Zool. Soc. 
London, p. 412: Nairobi; Longido West; Morogoro. Barbour & Loveridge, 
1928, Mem. Mus. Comp. ZooL, 50, p. 194: localities in Uluguru and 
Usambara Mountains. Loveridge, 1929, U. S. Nat. Mus. Bull. 151, p. 98: 
localities in Kenya and Tanganyika. 

Rana delalandii (part) Nieden, 1915, Mitt. Zool. Mus. Berlin, p. 352: Kiliman- 
jaro to Nguru; Amani; Ukami; Mpwapwa. 

Rana chapini Noble, 1924, Bull. Am. Mus. Nat. Hist., 49, p. 214, text fig. 6a: 
Batama, Belgian Congo. 

8 (M. C. Z. 16726-33) Mpwapwa, Ugogo. 23. xi. 29. 

Distribution. Already, under the name of delalandii, recorded from 
Mpwapwa by Nieden. 

The apparently discontinuous records of this frog will probably be 
connected by search in L^ganda and on Mt. Elgon or by reexamination 
of material from this region which has been identified as fuscigula and 
angolensis or nutti. Though it is the rain forest representative of 
fuscigula, in the highlands of Kenya and the Usambara Mountains 
chapini occurs in mountain streams apart from forest. 



362 bulletin: museum of comparative zoology 

Variation. An Uluguru frog was submitted to Dr. G. K. Xoble for 
favor of comparison with the type of chapini; in his reply he stated 
that he could detect no differences and considered them specifically 
identical. For further discussion on their relationships see R. f. 
angoIen.sis, of which nutii Boulenger is a synonym. 

Measurements. All eight specimens are juveniles just out of the 
tadpole stage, but fropi the data furnished below it will be seen that 
this form is the largest of all the races offuscigula. 

Noble's holotype cf 78 mm. (Belgian Congo.) 

From a total of 25 frogs, Procter records the largest 9 83 mm. Of 
193 frogs Barbour & Loveridge record d^ 74 mm., 9 110 mm. (Ulu- 
guru). Of 148 Loveridge gives maximums as cf 65 mm., 9 95 mm. 
(Kenya). 

I think that the reason for so few collectors obtaining examples of 
large size is to be attributed to the fact that the very big frogs keep to 
the swift-flowing streams. In consequence they are more difficult to 
catch as well as being harder to hold, so that natives, when bringing 
in frogs, are apt to confine their attention to the smaller indi\iduals. 

Rana fuscigula angolensis Bocage 

Rana angolensis Bocage, 1866, Jorn. Sci. Math. Nat. Phys. Lisboa, p. 73: 
Duque de Braganga, Angola. Parker, 1931, Proc. Zool. Soc. London, 1930, 
p. 897: Amatongas, Mozambique. 

Rana nutti Boulenger, 1896, Ann. Mag. Nat. Hist. (8), 18, p. 467: Lake Tan- 
ganyika. 

Rana delalandii (part) Nieden, 1915, Mitt. Zool. Mus. Berlin, p. 352: Ikombe; 
Kidugallo (Kidugala); Ubena; Ujiji; West Ruanda. 

21 (M. C. Z. 16520-9) Dabaga, Uzungwe Mtns. 1. i. 30. 
1 (M. C. Z. 16530) Panga Mawe, Uzungwe Mtns. 8. i. 30. 
Spawn, tadpoles & 9 (M. C. Z. 16531-9) Kigogo, Uzung^ve Mtns. 13. i. 30. 
1 (M. C. Z. 16540) Lukungu, Ubena Mtns. 8. ii. 30. 

1 (M. C. Z. 16541) Mangoto, UKinga Mtns. 10. ii. 30. 

12 (M. C. Z. 16542-51) Madehani, Ukinga Mtns. 10. ii. 30. 

47 (M. C. Z. 16552-61) Ilolo, Rungwe. 15. iii. 30. 

25 (M. C. Z. 16562-9) Nyamwanga, Poroto Mtns. 17. iii. 30. 

3 (M. C. Z. 16570-2) Nkuka Forest, Rungwe Mtn.8. iv.30. 

Tadpoles & 2 (M. C. Z. 16573-5, 17144) Tukuyu, Rungwe. 21. iv. 30. 

2 (M. C. Z. 17145-6) Ujiji, Lake Tanganyika. 25. v. 30. 

Distribution. Already under the earlier, but preoccupied, name of 
delalandii Dumeril & Bibron, Nieden has recorded this frog from 
Ubena and Ujiji. 



loveridge: African herpetology 363 

The Museum of Comparative Zoology possesses examples of this 
race from localities in the Cape Province and Orange Free State in 
the Union of South Africa; from Bella Vista, Angola; Waterfal Onder 
and Woodbush Village in the Transvaal; Behungi Escarpment, 
Uganda. 

I have also examined large series in the Field Museum of Natural 
History from various localities on Mt. Ruwenzori, arid from Lake 
Bunyoni and the Kigezi district of Uganda. 

Naiive names. Miula (Kihehe); chula (Kikinga). 

Afflniiics. Hewitt now considers that angolcnsis should be regarded 
as a race oi fuscigida and I wholly concur with his action; such treat- 
ment offers a reasonable explanation of the apparent intergrades 
along the boundaries of the two forms. 

Xieden, De Witte and others have long considered mdti Boulenger 
as a synonym, and it was in the hope of throwing more light on the 
much discussed relations of the group that I collected the above series. 
I am now convinced that nutti is a synonym. 

The synonymy is very involved for a tangle arose as a result of 
Boulenger failing to differentiate a third form — chayini Noble and 
thus, using specimens of chapini which Boulenger had identified as 
nidii. I continued to use the name but have been in reality really 
referring to chapini. 6ther authors have likewise been talking at cross 
purposes. 

More recently Parker (1931, loc. rit. p. 897) has suggested means for 
differentiating angolensis and nutti and provides a key. I imagine that 
his material was insufficient for in endeavoring to utilize this key I 
found all the characters very variable, the variations being largely 
correlated with age but by no means always so. 

As the holotype of nutti could not be loaned, Mr. Parker kindly sent 
me one of the Ruwenzori frogs identified as 7iutti by Boulenger (1909, 
Trans. Zool. Soc. London, p. 240, pi. viii, figs. 1 and 2) and it was at 
once obvious that it w^as distinct from the Nairobi frogs ( = chapini) 
identified for me as nutti by Mr. Boulenger in 1915. I then borrowed 
twenty -two Ruwenzori frogs from the Field Museum of Natural His- 
tory; these are undoubtedly conspecific with Boulenger's Ruwenzori 
frogs and agree well with his excellent figures. The data from them 
may be arranged as follows. 

In this Ruwenzori series of 22 frogs I found that the snout is from 
1 1/5 (49 mm. frog) to 1 3/5 (27 mm. frog) the diameter of the orbit; 
the nostril is nearer the eye than to the end of the snout in 10 frogs, 
equidistant in 12; the tympanum varies from 3/2 (27 to 36 mm. frogs) 



364 bulletin: museum of comparative zoology 

to 2/3 (38 to 67 mm. frogs) the diameter of the eye; the length of the 
foot is contained from 1 1/3 (48 to 50 mm. frogs) to 1 3/5 (67 mm. frog) 
times in the length from snout to anus. 

As an independent cheek I asked a student, Mr. J. B. White, if he 
would be so kind as to measure up two lots in the present series. I am 
obliged to him for taking the following measurements and working 
out the proportions. 

In a Dabaga series of 17 frogs the snout is from 7/8 (22 mm. frog) 
to 1 4/5 (53 mm. frog) the diameter of the orbit; the nostril is nearer 
the eye than to the end of the snout in 11 frogs, equidistant in 3, and 
nearer the snout in 3; the tympanum varies from Yl (26 mm. frog) to 
10/11 (53 mm. frog) the diameter of the eye; the length of the foot 
is contained from 1 3/5 (48 to 51 mm. frogs) to 2 (24 to 43 mm. frogs) 
times in the length from snout to anus. 

In a Nyamwanga series of 17 frogs the snout is from 1 1/3 (24 mm. 
frog) to 1 5/6 (50 mm. frog) the diameter of the orbit; the nostril is 
nearer the eye than to the end of the snout in 6 frogs, equidistant in 5, 
and nearer the snout in 6; the tympanum varies from 3<4 (46 mm. frog) 
to 2/3 (43 mm. frog) the diameter of the eye; the length of the foot is 
contained from XYi (43 mm. frog) to 1 4/5 (25 to 74 mm. frogs) times 
in the length from snout to anus. 

From this it will be seen that if a sufficient series be taken from 
any locality within its range, considerable instability is to be observed. 
An apparent increase in leg length as one proceeds from north to 
south may possibly be attributable to a disproportion of the sexes for 
the whole series of 1930 shows 57 in which the tibio-tarsal articulation 
falls short of the end of the snout, in most cases reaching the nostril, 
and 59 in which the tibio-tarsal articulation reaches beyond the end 
of the snout. There are two phalanges of the fourth toe constantly 
free of web. The whole series, with one exception, possess a tibia 
which is more than half the length from snout to anus ; the single ex- 
ception from Lukungu does not appear to differ in other characters 
from the rest of the series so I prefer to regard it as an abnormal 
f us cifiula -like individual. 

The three forms may be recognized as follows : 

Length of tibia not more than half the length from 

snout to anus ; fifth toe webbed to the very tip f.fuscigula 

Length of tibia more than half the length from snout 

to anus ; fifth toe webbed to the very tip /. chapini 

Length of tibia more than half the length from snout 

to anus, last phalanx of fifth toe free of web /. angolensis 



loveridge: African herpetology 365 

Coloration in life. Dabaga. Above, silvery -green or bronze to dark 
brown, an interorbital streak usually visible, a broad black patch from 
eye over tympanum; the whole upper surface is mottled with brown; 
the hind limbs are more or less distinctly cross-barred; no anal streak. 
Below silvery -white to cream; the lower jaws, throat and belly marbled 
with brown; the underside of the limbs opaquely greenish white. 

In the Kigogo frogs the gular and abdominal vermiculations were 
less distinct but these frogs were taken some distance from water. The 
vertebral stripe of one Ilolo frog was bright vermilion, the only one 
seen of such a color. 

Measurements. A quite exceptionally large 9 , the largest of the 
whole series measures 74 mm. 

Breeding. Eggs, tadpoles and young frogs with stumpy tails were 
taken in a fast-flowing stream at Kigogo on 13. i. 30; tadpoles were 
found at Tukuyu on 21. iv. 30 and frogs with rudimentary tails at 
Ujiji on 25. v. 30. 

Parasites. Parasites were common about the anus and posterior 
surface of the thighs in frogs from Dabaga and elsewhere. 

Enemies. A young frog was recovered from the stomach of an 
immature Striped Schaapsteker {Trimerorhinus f. tritaeniatus). 

Habitat. At Dabaga these handsome frogs rest in tussocks of grass 
beside the swiftly flowing brook into which they leap as one ap- 
proaches. Some dive to the bottom, others come up among the 
numerous reeds and grasses growing from the water. While usually 
associated with the swiftly -flowing brooks in valley bottoms, atMade- 
hani they were often encountered away from water in long sodden 
grass, or among the leaves of the bamboo forest; an even more sur- 
prising situation was among the leaves on the floor of the dark rain 
forest, a situation in which I captured several in the depths of the 
Nkuku Forest on Rungwe. 

Folklore. The first Kigogo frog was brought by an IVIbena lad who 
had tied it between two sticks, I told him to hold it by the hind leg and 
he replied, "No, no, I'm afraid of it." However, I told him not to be 
silly whereupon he took it from me and carried it to camp. I found a 
revulsion to handling frogs quite common among the Wabena though 
without ascertaining the cause. 



Rana galamensis Dumeril & Bibron 

Rana galamensis Dumeril & Bibron, 1841, Erpet. Gen., 8, p. 367: Galam Lakes, 
Senegal. 



366 bulletin: museum of comparative zoology 

Rana bravanus Tornier, 1897, Kriechthiere Deutsch-Ost-Afrikas, p. 92, fig. f.: 

Bagamoyo, etc. 
Rana bravana Nieden, 1915, Mitt. Zool. Mus. Berlin, 7, p. 351: Localities in 

Tanganyika Territory. 

Tadpoles and 3 (M. C. Z. 16576-8) Bagamoyo. 16. xi. 29. 

1 (M. C. Z. 16579) Ukerewe Id.. Lake Victoria. 14. vi. 30. 

Disiribidion. I had previously collected this frog at Bagamoyo and 
Bukoba; Nieden has recorded it from Pemba and Zanzibar. 

Affiniiies. Despite Nieden's contention that bravana should be con- 
sidered distinct from galamensis, I adhere to Boulenger's reiterated 
opinion that the two are synonymous; presumably the point will not 
be definitely settled until fi-esh material from the Galam lakes is 
available. 

Variation. It seems very doubtful if Boulenger's R. darlingi from 
Mashonaland is more than a race, or even distinct; in 1910 the only 
points on which he could separate it were that the vomerine teeth were 
behind the level of the choanae (between in (lalamensis) and the tibio- 
tarsal articulation of the adpressed hind limb reaching between the 
eye and the tip of the snout (not reaching beyond the eye in gaknn crisis). 
In the eight examples from Zanzibar and Tanganyika Territory in the 
collection of the Museum of Comparative Zoology it reaches from the 
hind end of the eye to between the eye and nostril ; the vomerine teeth 
are certainly between the choanae in this series but they are very 
variable in size and exact location. Tornier (1897) has figured some of 
these variations. 

Measurements. All adult, the males measure 67 to 71 mm., and the 
females 73 to 75 mm. 

Coloration. The white streak on the lips and the brown lateral 
stripes make this species easy to distinguish. The males have black 
vocal pouches. 

Breeding. Tadpoles collected on November 11th at Bagamoyo may 
possibly l)e referable to R. adspersa. They swarmed in seething 
masses or shoals which set the water in commotion, the patches of 
tadpoles cover areas of a foot to eighteen inches in length by half as 
much in width. 

Habitat. This is the most secretive of any of the East African ranae; 
it keeps to grass-grown, deep-water swamps apparently and though 
noisy enough, dives at the slightest disturbance in its vicinity. It is 
little wonder that, despite its wide distribution, it is scarce in collec- 
tions. After vainly trying for an hour to locate some in a swamp, one 
was observed on damp ground at the edge of a pool measuring twenty 



loveridge: African herpetology 367 

feet long by fifteen across but onh' knee-deep in the centre. Clumps 
of rushes grew in the shallow edges of the pool, rank sedges or papyrus 
almost to the edge. I had all this cut down and cleared away till the 
place was bare for Salimu said that he had heard two frogs calling 
"meow-meow" in this pool. We caught the larger frog seen on land 
which proved to be a female and after much waiting we secured two 
males and also an ads per sa. They call louder and continuously when 
the sun shines but become silent, except for an occasional call, when it 
rains. 

Rana floweri Boulenger 

Ranafloweri Boulenger, 1917, Ann. Mag. Nat. Hist. (8), 20, p. 417: Rosaires, 

Blue Nile. Type a 9 . 
Rana barbouri Loveridge, 1925, Proc. Zool. Soc. London, p. 776: Nj'ambita, 

Mwanza, Tanganyika Territory. Type a 9 . 
Abrana cotti Parker, 1931, Proc. Zool. Soc. London, p. 898: text fig. 1, Charre, 

Mozambique. Tj^pe a 9 . 

24 (M. C. Z. 16580-99) Mangasini, Usandawi. 14. xii. 29. 
1 (M. C. Z. 16600) Shinyanga, Usukuma. 3. vi. 30. 

Distribution. It might be remarked that Shinyanga is not very far 
south of Xyambita, the type locality of barbouri. This frog, super- 
ficially so like Ra)ia oxyrhynchus but with shorter hind limbs, evidently 
ranges from the Sudan to Mozambique. 

Affinities. Boulenger described floweri on the basis of a single 
female which he considered closel}' related to the subgenus Ptychadena; 
when I described barbouri I relegated it to the subgenus liildebrandtia 
being in error in supposing it had a hair-like clavicle. Parker points 
out that his eight specimens lack clavicles and proposes a genus for 
them which he calls Abrana. It is a matter of personal opinion whether 
Abrana merits full generic rank; to me it appears a mistake to mask a 
frog of so distinctly a ranid appearance by erecting a new genus for its 
reception, I prefer to consider Abrana a subgenus of Rana. Mr. Parker 
has kindly examined floweri and fully concurs that cotti is synonymous 
but would retain Abrana as a full genus. I have compared the holotype 
of barbouri with a paratype of cotti. Other examples of this frog in 
the collection of the Museum of Comparative Zoology are from Giza, 
Egypt and Frere Town, Kenya Colony. 

Variations. Parker has already pointed out some interesting varia- 
tions in his paratypes; if one compiles a new description based on the 
three earlier descriptions it will be found to cover most of the varia- 



368 bulletin: museum of comparative zoology 

tions. Boulenger speaks of a small flat outer metatarsal tubercle while 
Parker and myself say that there is no outer tubercle. There is no 
trace of it in the type of barhuuri and it is absent in nine males and 
three females of the Mangasini series but quite distinct in nine males 
and one female from the same locality, it is large in the two young of 
33 and 34 mm. The tibio-tarsal articulation marks the tympanum in 
two frogs, the posterior border of the eye in ten, the anterior border of 
the eye in eleven and beyond that in one (M. C. Z. 16599). Though 
apparently conflicting it is equally true to say that the snout is pointed 
and abruptly truncated at the tip. In the field I made a note to the 
effect that the dorsal and dorso-lateral folds almost entirely disappear 
when these frogs are immersed in water for from 24 to 4S hours; this 
is the condition in the type of harhouri; Parker has also remarked on 
the inconstancy of these glandular folds in his series. 

Measurements. Eighteen males range from 42 to 49 mm., average 
45.5 mm.; three females are from 43 to 48 mm., average 46.2 mm.; 
young are 33 and 34 mm. 

Breeding. The males were assembling after the first downpour of 
the rains and vast numbers were pairing in the water of the flooded 
flats. Males were calling and apparently sometimes inflated the 
abdominal skin instead of the singing pouches. Whatever the cause, 
accidents occurred in which the abdominal skin becomes inflated, this 
causes the frog to turn over on to its back when it flounders help- 
lessly; a round, white, globular skin the size of a ping-pong ball being 
all that is to be obser\ed floating on the water at a short distance. 
Two frogs were caught in this condition. 

Diet. vSpiders in one examined. 

Habitat. While the Mangasini frogs were found as described in the 
note on breeding above, the Shinyanga specimen was captured in my 
bedroom where it was jumping against the walls and had been doing 
so apparently for some time as the terminal joints Avere already worn 
off its toes. 



Rana oxyrhynchus Smith 

Rana oxyrhynchus A. Smith, 1849, Illus. Zool. S. Africa, 3, pi. Ixxvii, figs. 2, 

2a-c: Kafirland and region of Port Natal. 
Rana theileri Mocquard, 1906, Bull. Mus. d'Hist. Nat. Paris, p. 252: Nelspruit, 

Transvaal. 

8 (M. C. Z. 16601-8) Bagamoyo. 9. xi. 29. 

8 (M. C. Z. 16618-25) Kipili, Lake Tanganyika. 19. v. 30. 



loveridge: africax herpetology 369 

Distribution. Nieden has recorded this species from Dar es Salaam, 
Mpwapwa, Tukiiyu and I jiji. 

.iffinitics. There is nothing in the description of R. theilcri to lead 
one to suspect that this frog differs from R. o.vyrhynchos and I am 
indebted to Mr. V. FitzSimons for confirming this opinion after ex- 
amining the series of oxyrhynchus from Xelspruit in the Transvaal 
Museum. 

Variatiun. Mostly young specimens in which the hind limbs appear 
rather short, the body is included in the total length of the hind limb 
1.5 to 1.9 times in the Bagamoyo series, 1.7 to 2.1 times in the Kipili 
frogs, the averages being 1.7 and 1.9 respectively. 

Coloration in life. See habitat. 

Measurements. Largest male measures 40 mm.; the largest female 
45 mm., both are from Bagamoyo; average length of the Bagamoyo 
frogs is 30 mm., and of those from Kipili 31 mm. 

Habitat. At Bagamoyo they were taken in water holes on the sea 
front Avhere they exhibited great variety in color and markings. At 
Kipili scores of these frogs, grey to sandy in color, occasionally pink 
or still more rarely with a green vertebral stripe, were resting on the 
damp sand within five feet of a stagnant little lagoon very close to the 
edge of the lake. As one walked along they leaped towards the water 
in astonishing numbers, it was like walking through a swarm of locusts. 

Rana mascareniensis mascarexiensis Dumeril & Bibron 

Rana mascareniensis Dumeril & Bibron, 1841, Erpet. Gen., 8, p. 350: Madagas- 
car; Mauritius; Seychelles. 

9 (M. C. Z. 16609-17) Mwaya, Lake Nyasa. 1-8. iii. 30. 
3 (M. C. Z. 16640-2) Tukuyu, Rungwe. 13. iii. 30. 

3 (M. C. Z. 16643-5) llolo, Rungwe. 15. iii. 30. 

151 (M. C. Z. 16646-57) Nyamkolo, Lake Tanganyika. 9. v. 30. 

4 (M. C. Z. 17826-9) Kitungulu, Urungu. 14. v. 30. 

1 (M. C. Z. 16658) Kasanga, Lake Tanganyika. 16. v. 30. 
12 (M. C. Z. 16659-68) Ujiji, Lake Tanganyika. 28. v. 30. 
3 (M. C. Z. 16669-71) Ukerewe Id., Lake Victoria. 14. vi. 30. 

Distribution. Recorded from Bagamoyo, Dar es Salaam, Tukuyu 
and Ukerewe Island by Xieden. 

Xative name. Makeri (Kijiji). 

Variation. The Mwaya series was measured for head into hind limb 
length and was found to be 1.7 to 1.9, i.e. within the range of R. 
oxyrhynchus. 

Measurements. Omitting the 160 young under 33 mm., the fifteen 



370 



bulletin: museum of comparative zoology 



males measure from 33 to 46 mm., average 39 mm.; the eleven females 
measure from 33 to 53 mm., average 44 mm. 

Breeding. Great numbers were calling "quek-quek" in the swamp 
eight miles west of Bagamoyo town. The Mwaya female is distended 
with ova. At Tukuyu three males were taken as they were calling 
from a shallow pool, a clear, ringing, liquid note something suggestive 
of a bursting bubble and "bob-white." When caught and placed in 
a bag they only gave utterance to the "quek-quek" cry. At Nyamkolo 
scarcely any adults were seen but young were present in thousands in 
the swamped grasslands at the edge of the lake where their green 
vertebral stripe might serve a useful purpose. At dusk they leave the 
swamped grass for the adjacent meadowland. 

Enemies. Recovered from the stomach of a Cattle Egret {Bubul- 
culus ibis) at Bagamoyo, from a Hissing Sand Snake {Psammophis 
sihilans) at Nyamkolo, from a Lined Snake {Dromophis lineatus) at 
Ujiji, and it was apparently this species of frog which was found in a 
water snake (Grayia thoUoni) on Ukerewe Island. 

Habitat. Abundant in swamps or on the lake shore at Mwaya, 
Nyamkolo and Kasanga. 

Rana mascareniensis uzungwensis Loveridge 

Rana mascareniensis uzungwensis Loveridge, 1932, Bull. Mus. Comp. Zooi., 
72, p. 384: Dabaga, Uzungwe Mtns., Tanganyika Territory. 

12 (M. C. Z. 16626-35) Dabaga, Uzungwe Mtns. 1. i. 30. 
3 (M. C. Z. 16636-8) Kigogo, Uzungwe Mtns. 13. i. 30. 
1 (M. C. Z. 16639) Lukungu, Ubena Mtns. 8. ii. 30. 

Breeditig. Males were assembling in a patch of boggy land at 
Dabaga where only one female, the type, Avas taken, this specimen 
was full of ova. At Kigogo a young female measuring 2S mm. was 
captured. 

Habitat. The type series was taken in boggy land bordering a 
swiftly flowing brook in the bottom of the valley on the left side of the 
road as one approaches Mr. A. K. Hauter's farm from the direction of 
Iringa. 



Rana mascareniensis venusta Werner 

Rana venusta Werner. 1907, Sitzber. Akad. Wiss. Wien, 116, part 1, pp. 1889, 
and 1892, pi. iv. fig. 11: Entebbe, Uganda; Mongalla and Lagos. 

4 (M. C. Z. 16672-5) Entebbe, Lake Victoria. 27. vi. 30. 



loveridge: African herpetology 371 

Distribution. This is the large western form of mascareniensis, its 
(hstribution more or less coinciding with that of the rain forest. The 
above specimens are topotypes. 

Measurements. Only the male is fully grown, measuring 49 mm., 
the larger young female measures 40 mm. 

Variation. The width of the interorbital space in its relation to the 
width of an upper eyelid proves valueless for distinguishing this race 
from the typical form. Nor are the markings on the buttocks constant 
though the light line is usually present. The race, however, may be 
told by the difference in webbing of the toes and its larger size, females 
attaining a length of 63 mm. 

Rana ansorgii Boulenger 

Rana ansorgii Boulenger, 1905, Ann. Mag. Nat. Hist. (7), 16, p. 107, pi. iv, 
fig. 1: Between Benguella and Bihe, Angola; Parker, 1931 (1930), Proc. 
Zool. Soc. London, p. 898: Amatongas, Mozambique and Sibundeni, 
Zululand. 

12 (M. C. Z. 16676-85) Kitungulu, Urungu. 14. v. 30. 

Distribution. These examples constitute the first record of the 
occurrence of this species in Tanganyika Territory. 

Affinities. Two frogs from this series were submitted to Mr. H. W. 
Parker for favor of comparison with the type, w^ith the figure and 
description of which they appeared to be in accord. Mr. Parker 
replied: ''These two agree perfectly with specimens from the Victoria 
Falls, determined by Boulenger as R. ansorgii. I have also compared 
them with the type and find very close agreement." 

The species seems to represent a still further development beyond 
R. m. uzungwensis in the direction of the Rana fasciata group for in 
ansorgii the 1st, 2nd, 3rd and 5th toes have two joints completely 
free of web as against one, or one and a half, in mascareniensis, the 
4th toe in the latter has two joints free as against three free joints in 
ansorgii. The tibio-tarsal articulation of the adpressed hind limb 
extends far beyond the tip of the snout. 

Coloration in life. These frogs were of a bright straw shade similar 
to that of Rana fasciata merumontana. 

Measurements. All are females and range from 23 to 31 mm. in 
length. 

Habitat. The series was found in swamped forest land where most 
of the trees had been recently felled, the spot was only fifty feet from 
a swiftly flowing stream. 



372 bulletin: museum of comparative zoology 



Rana fasciata merumontana Lonnberg 

Rana merumontana Lonnberg, 1907, in 1910, in Sjostedt, Kilimandjaro-Meru 
Exped., 1, pt. 4, p. 21, pi. i, figs. 4a & 4b: Mt. Meru, Tanganyika Territory. 

Rana fulleborni Nieden, 1910, Sitzber. Ges. Naturf. Freunde Berlin, p. 436: 
Crater Lake of Ngosi Volcano, Tanganyika Territory. 

Rana fasciata merumontana Barbour & Loveridge, 1928, Mem. Mus. Comp. 
Zool., 50, p. 197: Phillipshof, Usambara Mtns., Tanganyika Territory. 

4 (M. C. Z. 16686-9) Dabaga, Uzungwe Mtns. 1. i. 30. 
2 (M. C. Z. 16690-1) Kigogo, Uzungwe Mtns. 13. i. 30. 
1 (M. C. Z. 16692) Mangoto, Ukinga Mtns. 10. ii. 30. 
4 (M. C. Z. 16693-6) Madehani, Ukinga Mtns. 19. ii. 30. 
58 (M. C. Z. 16697-725) Nyamwanga, Poroto Mtns. 17. iii. 30. 

Distribution. The East African race of the South African and 
Angolan R.f. fasciata is only known from the localities in the citations 
above and the recently collected material of which the big series 
from the Poroto Movmtains are practically topotypes, one at least 
coming from the slope of Ngosi Volcano. 

Native names. Jcraboka (Kihehe); c/z?//o (Kikinga). 

Variation. Rana fulleborni is something of an intermediate between 
the typical form and merumontana. Unfortunately our South African 
material has not been preserved in the same way as the East African 
which makes comparisons rather difficult. The sharply distinct dorso- 
lateral ridges or folds oi fasciata which tend to be broken up in meru- 
montana may owe their distinctness in part to preservation in strong 
alcohol, however that may be the topotypes oi fUUcborni cannot be 
distinguished from Usambara specimens of merumontana on this 
character though the dorso-lateral folds average a better development 
in frogs from the Poroto Mountains. 

Another supposedly distinguishing character is the interorbital 
width in its relation to that of an upper eyelid, twice the width of an 
upper eyelid in merumontana, equal to it in filUeborni. Each frog of 
the Poroto series was examined as soon as chloroformed and a note 
made that in all the interorbital space equalled an upper eyelid in 
width; after a year, first in formalin, then in alcohol, however, there 
is considerable variation from 1 to 1^ times; certainly they average 
the same in this respect as the big series from the Usambara Moun- 
tains. There is no difference in the position of the nostril which agrees 
with the variation given in the 1928 citation above. 

One frog which lacks a right hind leg appears to have never de- 
veloped it. 



loveridge: African herpetology 373 

Coloration in life. Daha<i;a. Above, straw-brown, a conspicuous 
yellowish green vertebral streak bordered on either side by a black 
raised skin fold extending from snout to anus, a more straw-colored 
streak edged above and below with black, commences at the eye and 
merges into the abdominal coloring in front of the groin, below this 
anteriorly is a broad dark brown streak from eye to fore limb which is 
buft', mottled with brown; an anterior and a posterior streak on the 
tibia continued on foot and for a short distance on the thigh, the dorsal 
area between these streaks buff, faintly mottled or marbled with 
brown. Below, cream, very inconspicuously sprinkled with brown on 
the throat, undersides of the limbs colorless, the bones more or less 
sharply distinguishable through the skin. 

A young one taken in boggy ground at the foot of Ngosi Volcano 
was bright yellow like the local Hyperolius marginatus which were 
common in the same locality, and strikingly different from the adults 
of its own species. The throats of the Nyamwanga series are slightly 
dusky, even freckled in some of the males as in the type oi fulleborni. 

Measurements. The largest specimens from all localities were 
females and measured 50 mm., which would appear to be the maxi- 
mum size as the largest of a hundred and six Usambara frogs was 46 
mm. 

Breeding. A Dabaga female is distended with ova. 

Diet. Beetles, grasshoppers and a walking-stick insect. 

Habitat. To the notes already furnished on the Usambara frogs one 
might add that at Dabaga, Mangoto and Nyamwanga our series were 
taken in the long lush grass of the valleys, these situations were 
generally near streams. At Kigogo, however, they were found on a 
hill top far from water, the long grass in which they lived was ap- 
parently very dry but at its roots there was shade and moisture. 

Phrynobatrachus natalensis (Smith) 

Stenorhynchus natalensis A. Smith, 1849, lUus. Zool. S. Africa, 3, Appendix, 
p. 24: Port Natal. 

1 (M. C. Z. 16893) Mpwapwa, Ugogo. 22. xi. 29. 

2 (M. C. Z. 16894-5) Mufindi-Njombe Road. 6. ii. 30. 
1 (M. C. Z. 16896) Below Senjeri Pass. 5. v. 30. 

1 (M. C. Z. 16897) Ukerewe Id., Lake Victoria. 14. vi. 30. 

2 (M. C. Z. 16898-9) Kampala, Uganda, vi. 30. 

Distribution. Nieden records this species from Zanzibar, Mpwapwa 
and Unyika. 



374 bulletin: museum of comparative zoology 

Coloration in life. The male from Mufindi-Njombe Road had a very 
broad, pale brown, vertebral band; most unusual in this species. 

Measurements. The four males range from 19-31 mm., average 
26 mm.; the three females 30-34 mm., average 32 mm. 

Breeding. Calling on February 6th from rain-filled ruts. 

Habitat. The pair from near Mufindi, as well as the frog from the 
Senjeri Pass, were taken from rain-filled ruts in the road. 



Phrynobatrachus acridoides (Cope) 

Staurois acridoides Cope, 1867, Journ. Acad. Nat. Sci. Philad., 6, p. 198: 
Zanzibar. 

Skeleton and 3 (M. C. Z. 16886-9) Miritini, Kenya Colony. 30. x. 29. 

3 (M. C. Z. 16890-2) Changamwe, nr. Mombasa. 31. x. 29. 
3 (M. C. Z. 16851-3) Mpoponi, Zanzibar. 21. x. 29. 
21 (M. C. Z. 16854-63) Bagamoyo, T. T. 9. xi. 29. 
3 (M. C. Z. 16864-6) Unyanganyi, Turu. 5. xii. 29. 
1 (M. C. Z. 16867) Handa, Usandawi. 10. xii. 29. 
1 (M. C. Z. 16868) Mwaya, Lake Nyasa. 1-8. iii. 30. 

1 (M. C. Z. 16869) Kasanga, Lake Tanganyika. 16. v. 30. 
5 (M. C. Z. 16870-4) Kipili, Lake Tanganyika. 19. v. 30. 

2 (M. C. Z. 16875-6) Shinyanga, Usukuma. 3. vi. 30. 

17 (M. C. Z. 16877-85) Ukerewe Id., Lake Victoria. 14. vi. 30. 

Distribution. Already recorded from Zanzibar, Bagamoyo, Dar es 
Salaam, Mpwapwa, Iringa and Tukuyu by Nieden. 

Native name. Koko (Kinyakusa) . 

Variation. I feel confident that digital disks are a breeding season 
development, they are present only in the Zanzibar, Bagamoyo and 
Mwaya frogs; that they are not present in the ]\Iiritini and Chan- 
gamwe specimens may be attributed to the low grade local spirit in 
which, as an emergency measure, I had to preserve specimens from 
these places. 

Coloration in life. While the majority of the Miritini frogs were 
brown, quite a number were bright green; on placing them in a bag, 
however, they all changed to brown. At Changamwe also, both green 
and brown frogs were present. The Kipili frogs were sand colored in 
which they resembled those from Dar es Salaam mentioned in the 
1928 report. 

Breeding. On October 30th at INIiritini a torrential downpour of 
rain formed pools in the grass beside the railway line, actually there 
had been several showers daily since the 27th inst. A chorus of calls 



loveridge: africax herpetology 375 

came from the pools and, on close examination, I found these frogs 
abundant on the edge of one of the pools. 

At Changamwe I visited a pool on October 31st; at the spot where 
these frogs were calling and a few pairs were found in embrace it was 
about six inches deep. The beaten-down blades of grass which lay 
on the surface of the water, held a narrow edging of spawn along their 
length, on some the spawn had already begun to take definite shape. 
There were also some patches of spawn about eight inches in diameter 
floating on the surface of the water free of grass, this appeared to be 
caused by all, or almost all, the available grass being occupied. 

At Bagamoyo, on November 11th, frogs of this species were calling 
vociferously the sound being like a miniature rattle; these frogs were 
in the large swamps eight miles west of the town. They were also 
abundant in waterholes where tadpoles, either of this species or of 
Bona m. mascareniensis were numerous ; the water holes were in sandy 
soil, almost on the shore. 

The Mwaya frog is a female distended with ova. Most of the frogs 
from Ukerewe Island are young, the smallest (M. C. Z. 16877) being 
12 mm. in length from snout to anus. 

Enemies. The Mwaya frog was recovered from the stomach of an 
Eyebrowed Viper {J ^i per a superciliaris). 

Habitat. Unyanganyi frogs were taken in boggy ground in an 
mbugive; the Handa specimen from a water hole in the valley bottom; 
the Kasanga frog at the edge of the lake; at Kipili in pools in sandy 
flats close to the lake; at Shinyanga in big pools in an otherwise 
dried-up and sandy river bed; on Ukerewe Island they were found on 
the sand along the edge of a slow-flowing stream. 

Phrynobatrachus perpalmatus Boulenger 

Phrynobatrachus perpalmatus Boulenger, 1898, Proc. Zool. Soc. London, p. 479, 

pi. xxxviii, fig. 1: Lake Mweru, N. Rhodesia. 
Phrynobatrachus perpalmatus werneri Ahi, 1924, Zool. Anz., 60, p. 273: El 

Grassi etc., Sudan. 

44 (M. C. Z. 16900-25) Nyamkolo, Lake Tanganyika. 9. v. 30. 
2 (M. C. Z. 17142-3) Sumbwa, Lake Tanganyika. 20. v. 30. 

Distribution. The type locality of this species is exactly 130 miles 
due w^est of Nyamkolo; the Sumbwa specimens constitute the first 
record of the occurrence of this species in Tanganyika Territory. 

Affinities. In 1924 Ahl proposed the name werneri for Sudanese 
specimens which he alleged differed from the type in : 



376 bulletin: museum of comparative zoology 

(i) web on the 4th toe being somewhat shorter 

(ii) interorbital space somewhat broader than an upper eyeUd 

(iii) tympanum completely hidden 

(iv) 1st finger is obviously shorter than the 2nd. 
In his original description Boulenger stated : 

(i) toes entirely webbed 

(ii) interorbital space a little narrower than an upper eyelid 

(iii) tympanum feebly distinct 

(iv) 1st finger not extending quite so far as 2nd. 

Boulenger, when stating that the toes were entirely webbed, was 
not anticipating the fine distinctions of some later taxonomists, 
actually the 4th toes may not be webbed to the tip. In the Xyamkolo 
series the interorbital space may be much narrower or much broader; 
in fact M. C. Z. 16914 has it nearly twice as broad, has the tympanum 
indistinguishable on the right side of the head yet clearly distinct on 
the left, while the presumed difference in the matter of finger length 
betAveen Boulenger's and Ahl's frogs is purely imaginary. 

Breeding. The breeding season was ob\iously OAcr at Nyamkolo and 
the series ranges from adults to frogs (with unabsorbed tails) which 
measure S mm. in length from snout to anus. 

Diet. Minute beetles. 

Hahifat. Next to R. m. mascareniensis this species appears to be the 
most abundant in numljers of any frog at Nyamkolo. Both inhabit 
the same swamped grasslands interspersed with reeds which makes 
sweeping with a net difficult. They scoot along the surface of the 
water with great rapidity and are harder to catch than the Mascarene 
Frog, the variegated shades of green which is their livery tends to 
conceal them when they come to rest. They l)egin to call about an 
hour before sunset and continue till the rising sun grows too hot, or 
so it seemed. The combined noise of many hundreds is somewhat 
like the tink-tonk of tiny hammers beating on little anvils. When it 
can be isolated, however, it will Ije observed that a single cry is very 
metallic and reminds one of the individual notes of the musical instru- 
ment so commonly carried hy East African natives. 

Arthroleptis stenodactylus stenodactylus Pfeft'er 

Arthroleptis stenodactylus Pfeffer, 1892 (1893) Jahrb. Hamburg Wiss. Anst., 
10, p. 93: Kihengo, Tanganyika Territory; Barbour & Loveridge, part, 
1928, Mem. Mus. Comp. Zool., 50, p. 207: Dar es Salaam specimens only. 

3 (M. C. Z. 16926-8) Bagamoyo. 8. xi. 29. 



lovf:ridge: African herpetology 6n 

6 (M. C. Z. 16929-34) Mpwapwa, Ugogo. 23. xi. 29. 
1 (M. C. Z. 169357) nr. Ikombo, N. Rhodesia. 6. v. 30. 
9 (M. C. Z. 169358-43) Kitungulu. Urungu. 14. v. 30. 

Distribufion. Xieden's records, like those in the 192S citation above, 
appear to he composed of typical sfcNodacfylus and the mountain 
form which I have recently described under the name of ^4. s. vlugu- 
ruensis. 

Correction. On this last expedition an effort was made to get as 
much material of this group as possible to check the opinion which I 
held in 1928 that variahilis Matschie and ichytii Boulenger were 
synonyms of stenodaciylus. Now with almost topotypic material of 
stenodadylus and of ichytii I must reverse that opinion and recognize 
both whytii and variabilis as distinct. 

Of the material listed under stenodadylus in 192S only the Dar es 
Salaam specimens are really of the typical form all the material from 
the Uluguru and Usambara Mountains being of the mountain race. 

Atfinities. A. s. stenodactylus is the most specialized burrowing 
member of the group, its spade-like inner metatarsal tubercle is larger 
and its toes a trifle shorter than in ^4. whytii, it is doubtful whether 
the two can be distinguished except by actual comparison ; stenodacty- 
lus appears to have shorter limbs though this may be illusory rather 
than actual. 

Variation. In addition to the material listed above six other speci- 
mens from Dar es Salaam, Dutumi and Kilosa have been used. In 
these twenty-five frogs the tibio-tarsal articulation reaches the axilla 
only in three, the tympanum in fifteen, the eye in four; the 1st finger 
usually equals the 2nd but is sometimes a trifle longer or shorter; the 
tips of the toes are not swollen or dilated ; the shovel-shaped metatarsal 
tubercle is always longer than the inner toe; the length from the snout 
to anus is included in that of the hind limb from 1.12 to 1.5S times, 
average 1.33. 

Coloration in life. Kitungulu. Above, pinkish-brown, a fine light 
yellow vertebral line from the snout to the anus, just above the anus 
it is crossed at right angles by a similar light line extending right and 
left to the inner aspects of the knee joints, the chain-like dorsal mark- 
ings typical of the genus are usually present though sometimes absent, 
a dark streak commences at the nostril and passes through the eye to 
the tympanum; the lips are more or less regularly mottled with pale 
grey and white. Below, the throat, chest and anterior part of the belly 
are china-white, the underside of the limbs colorless, the soles of the 
hands and feet dusky. 



378 bulletin: museum of comparative zoology 

Measurements. The largest frog, a female from Mpwapwa, measures 
34 mm. 

Diet. Mostly termites, a beetle in one and what appeared to be an 
earthworm in another. 

Enemies. Three were recovered from the stomach of a Sharp- 
snouted Snake {Rhavvphiophis rostratus) at Dar es Salaam. 

Habitat. The Bagamoyo frogs were taken in sandy soil in the shade 
of a mango tree, in sandy soil at the base of a banana, and among 
leaves under a tree close to the seashore. All six Mpwapwa frogs were 
in holes among the decayed roots of a tree stump where they appeared 
to be aestivating as their stomachs were empty. The Ikombo and 
Kitungulu frogs in dry woodlands on red soil, an environment and 
climate closely resembling that at Kilosa. 

Arthroleptis whytii Boulenger 

Arthroleptis whytii Boiilenger, 1897, Proc. Zool. Soc. London, p. 802: Kondowe 
to Karonga, Nyika Plateau, Masnku Mountains. 

2 (M. C. Z. 16935-6) Mwaya, Lake Nyasa. 3. iii. 30. 
7 (M. C. Z. 16944-8) Outside Nkuka Forest, iii. 30. 
2 (M. C. Z. 16953-4) Tukuyu, Rungwe. 21. iv. 30. 

Affinities. These frogs are from very near the type locality as Mwaya 
is only thirty miles from Karonga; a comparison with Boulenger's 
figure of the type shows them to be specifically identical. The diffi- 
culty of separating this species from stenodactylus has been discussed 
under the latter. 

Variation. The tibio-tarsal articulation marks the eye; the inner 
metatarsal tubercle, though spade-like, is not so developed as in steno- 
dactylus but still is longer than the inner toe; the length from the snout 
to the anus is included in that of the hind limb from L13 to 1.63 times, 
average L36. 

Measurements. The largest frog, a male from the Nkuka Forest 
edge, measures 28 mm. 

Breeding. None were breeding and the small size of the Mwaya 
specimens indicates that the season had passed. 

Diet. The identifiable stomach contents consisted of: (1) a frog of 
apparently the same species and fi^'e beetle larvae, (2) spinose cater- 
pillar and a snail, (3 & 4) snails, (5) and (6) grasshoppers. 

Parasites. Female oxyuroid worms and parts of a proteocephalid 
cestode were present in one. 

Enemies. One of these frogs was recovered from the stomach of a 



loveridge: African herpetology 379 

Green Snake {Chlorophis hoplogasfer) at Ilolo, and apparently what 
was a young whytii in the stomach of a larger frog of this species. 

Habitat. The young were taken among leaves in wet woods at 
Mwaya, the adults were captured by myself in the wet roadw^ay just 
below the Xkuka Forest and belong to the plateau fauna. None was 
taken within the forest where most of our collecting was done; in the 
forest their place is taken by the closely related A . reichei. 

Arthroleptis adolfi-friederici Nieden 

Arthroleptis adolfi-friederici Nieden, 1910, Sitzber. Ges. Naturf. Freunde Berlin, 
p. 440, and 1912, Wiss. Ergebn. Deutsch-Zentral-Afrika-Exped., 4, p. 175, 
pi. V, figs. 4a-c: Rugege Forest, Belgian Ruanda. Barbour & Loveridge, 
1928, Mem. Mus. Comp. Zool. 50, p. 212: Localities in Uluguru and Usam- 
bara Mtns., Tanganyika Territory. 

1 (M. C. Z. 16952) Ngosi Crater, Poroto Mtns. 19. iii. 30. 

Distribution. This species has been recorded by Nieden as occurring 
on Rungwe. 

Variation. The tibio-tarsal articulation reaches far beyond the 
end of the snout. 

Measurements. This frog, a female, is 46 mm. in length or 4.5 mm. 
longer than the type. 

Habitat. Taken as it was hopping on the sodden forest floor within 
the crater. 

Arthroleptis reichei Nieden 

Arthroleptis reichei Nieden, 1910, Sitzber. Ges. Naturf. Freunde Berlin, p. 440: 
Crater Lake, Ngosi Volcano, Tanganyika Territory. 

1 (M. C. Z. 16949) Kigogo, Uzungwe Mtns. 23. i. 30. 

2 (M. C. Z. 16950-1) Madehani, Ukinga Mtns. 19. & 22. ii. 30. 

3 (M. C. Z. 16955-7) Ngosi Volcano, Poroto Mtns. 19. iii. 30. 
80 (M. C. Z. 16958-75) Nkuka Forest, Rungwe Mtn. iii. 30. 

Distribution. Hitherto only known from the type; the Ngosi Vol- 
cano series are topotypes. 

Native names. Buluwidi (Kikinga); koti (Kinyakusa). 

Affinities. Somewhat intermediate between ^1. u-hytii (from which 
it may be distinguished by the pronounced disks on the toes of reichei) 
and A. adolfi-friederici (from which it differs in possessing a first finger 
that is very much shorter than the second, and in its smaller size). 
In adolfi-friederici the first finger equals the second or is only a little 
shorter. 



380 bulletin: museum of comparative zoology 

Variation. There is little to add to the excellent description except 
that occasionally the tibio-tarsal articulation of the adpressed hind 
limb reaches as far as the end of the snout. 

Coloration in life. Xgosi Volcano. 25 mm. cf . Caught in a wild 
banana. Above, straw colored and very much like a young A. s. 
stenodactylus but soon changed to a greenish-grey; a dark '-^ shaped 
marking edged with white anteriorly, unites the upper eyelids; a 
sepia-brown line from the nostril crosses the edge of the eyelid and 
descends to, and behind, the tympanum where it terminates ; numerous 
sepia-brown, or black, light-edged spots on the jaws, sidts of head, 
body and limbs. Below, throat pluml)eus faintly marbled with white, 
the margins of the jaws flecked with pure white; rest of underparts 
greenish-yellow, tinged with orange on belly and groins; breast and 
sides of belh' vermiculated with brown and white. 

Measurements. The largest frog, a female from the Xkuka Forest, 
measures 32 mm. 

Breeding. Xoneofthespecimenstakenin March appear to be breeding. 

Diet. The food is generally too finely masticated to be recognisable, 
this was the case with seven frogs examined; in an eighth a bug, 
caterpillar and numerous small forest cockroaches were distinguishable. 

Enemies. A frog {Arfhroleptis reiehei) was recovered from the 
stomach of a Chlorophis nrglccfii.t at Kigogo, and another from a 
Crotaphoyeltis h. tornieri at Madehani. 

Habitat. One of the Madehani frogs was taken in a small patch of 
swampy ground caused by seepage from the mountain side; wild 
bananas were abundant in this damp spot. The Kigogo frog was in 
rain forest near the Ruaha River. As already indicated the Xgosi frogs 
were taken in wild bananas. Though these bananas were abundant 
in the ravines of the Xkuka Forest, we found no frogs in them, all our 
large series being taken on the leaf -strewn forest floor. They are ex- 
tremely quick to take cover, usually two hops and they have vanished, 
having slipped under the leaves or into the rotting vegetation, nor 
are they easily found once they have gained such a retreat. The 
species appears to be rare in its type locality but it should be re- 
membered that only three days were spent on the Xgosi Volcano as 
against three weeks in the Xkuka Forest. 

Arthroleptis schubotzi Xieden 

Arlhroleptis schubotzi Nieden, 1910, Sitzber. Ges. Naturf. Freunde Berlin, p. 
440; and 1912, Wiss. Ergebn. Deutsch-Zentral-Afrika-Exped., 4, p. 177, 
pi. V, fig. 3: Usumbura, Tanganyika Territory. 



loveridgk: africax herpetology 381 

2 (M. C. Z. 17026-7) Dabaga, Uzungvve Mtns. 1. i. 30. 
1 (M. C. Z. 17028) Kigogo, Uzungwe Mtns. 23. i. 30. 
1 (M. C. Z. 17140) Madehani, Ukinga Mtns. 14. ii. 30. 
349 (M. C. Z. 17029-53) Nkuka Forest, Rungwe Mtn. iii. 30. 

Distribution. Only known from the holotype anfl five other speci- 
mens recorded by Xieden from Mpwapwa and Tiikuyu. The frogs 
from Bagilo, L luguni Mountains recorded under this name (Barbour 
& Loveridge, 1928, Mem. Mus. Comp. Zool., 50, p. 213) appear to be 
only color variants of A. xenodactylus Boulenger. 

Native name. Koti (Kinyakusa, l)ut not specific). 

Affinities. Distinguished from its nearest East African allies by its 
very short hind limb, the tarso-metatarsal joint only reaching the 
eye in the type and in most specimens; in a very few individuals the 
tibio-tarsal articulation may reach the eye. 

Coloration in life. Madehani. 9 . Abo\e, l)luish-grey mottled with 
black, the specks being more or less arranged in longitudinal lir.es on 
the- back; thighs showing some light (blood) red, otherwise limbs 
colored as back. Below, very pale ))luish-white heavily marbled 
with darker, the reverse is the case on the limbs which are mainly 
dark with a few pale bluish specks; groin and a few patches on the 
tibia red. 

Measurements. The largest specimens, females, from Kigogo, 
Madehani and the Nkuku Forest all measure IS mm. 

Breeding. Undoubtedly breeding at Dabaga, Kigogo and Madehani 
where females, distended with eggs, were taken in swampy ground. 
At Madehani this was caused by seepage from the mountain side 
about fifty yards below the forest edge. In the same patch whicli was 
not more than five feet in diameter, two males and a female A. parvulus 
were taken together with a number of freshly hatched tadpoles still 
in the jelly. It is presumed that these were the larvae of parvulus. 
As about 300 of the Rungwe series consist of small young not as big 
as a house fly, it may be confidently assumed that the breeding season 
there coincides with that of the Uzungwe and Ukinga Mountains. 

Diet. A tick, a muscid fly, termites, ants and very small maggots 
were recognisable. The food of a frog, itself scarcely larger than a 
bluebottle, must necessarily be very small. 

Parasites. Rungwe frogs were commonly parasitised by a pink 
larval mite. 

Habitat. The Dabaga pair were taken six feet from the bank of a 
brook, A. minutus was present close by but in more marshy ground. 
The nearest surviving forest was several hundred yards away. I have 



382 bulletin: museum of comparative zoology 

little doubt that schuhoizi is common in the big forest where Colohus 
gordonorum occurs for I have recollections of seeing it on the sole 
occasion when I visited this forest in search of guerezas. 

The Kigogo frog was brought to me by Mr. H. Frazer who had found 
it in the forest nurseries abutting on a very small patch of rain forest. 

The first examples from Rungwe were five young, ranging from 8 
to 13 mm. in length, which I found among scraps of bark and chips 
from a felled tree in a clearing in the rain forest. Search revealed that 
the species was abundant, particularly in the numerous saw-pits 
scattered through the forest, these pits served as traps and from them 
the larger proportion of the series was obtained. 

Arthroleptis xenodactylus Boulenger 

Arthroleptis xenodactylus Boulenger, 1909, Ann. Mag. Nat. Hist. (8), 4, p. 496: 
Amani, Usambara Mountains, Tanganyika Territory; Barbour & Lover- 
idge, 1928, Mem. Mus. Comp. Zool., 50, p. 214: Uluguru and Usambara 
Mtns., Tanganyika Territory. 

Arthroleptis schubotzi Barbour & Loveridge {nee. Nieden), 1928, Mem. Mus. 
Comp. Zool., 50, p. 213: Bagilo, Uluguru Mtns., Tanganyika Territory. 

81 (M. C. Z. 16976-17000) Mpwapwa, Ugogo. 22. xi. 29. 

82 (M. C. Z. 17001-17025) Kitungulu, Urungu. 14. v. 30. 

Distribution. Known from the above localities and Kilosa. 

Variation. Compared with the topotype series in the Museum of 
Comparative Zoology with which they agree in possessing a single 
metatarsal tubercle, dilated finger tips and limbs of moderate length. 

Coloration in life. Extraordinary variation was to be seen in the 
Kitungulu series among which I noted the following principal types: 

(i) Above, brown tinged with pink, the whole upper surface ex- 
hibiting a metallic gloss; a yellowish vertebral line from the snout to 
a point just above the anus where it forks, almost at right angles, to 
terminate on the inner side of the knees; a dark triangular mark be- 
tween the eyes, its apex passing backwards and merging into two 
similar, but diamond-shaped, blotches, the whole forming the chain of 
markings characteristic of many members of the genus; a light line 
from the snout passes above the eye where it ceases, below it is a darker 
line extending backwards as far as the tj'mpanum. Below, colorless 
except the belly which is cream and the throat and flanks which are 
marbled with grey ; the hinder aspect of the thighs is profusely sprinkled 
with fine white spots. 

(ii) Above, bright red or reddish-brown, the vertebral line red; no 



loveridge: africax herpetology 383 

transverse line across the anal-thigh region; no liglit line from the 
nostril over the eye; otherwise the markings are as in (i). 

(iii) Above, bright yellowish-green, almost uniform, such markings 
as are present are so faint as to be indiscernible except upon close 
scrutiny with a lens. Below, exactly as in (i). 

Measurements. The largest frogs, females from Mpwapwa, measure 
21 mm., the largest from Kitungulu are 18 mm. 

Parasites. A female Oxyuroid worm was present in one frog ex- 
amined. 

Enemies. A frog of this species was found in the stomach of each of 
four young White-lipped Snakes {Crotapliopeltis h. hotamboeia) taken 
at Kitungulu. 

Habitat. The habitat is similar to that which I have recorded at 
Amani, viz. among an abundance of fallen leaves beneath trees. At 
both Mpwapwa and Kitungulu there was a stream in close proximity 
to the leaf -strewn ground upon which these frogs were found. At 
Kitungulu the leaves had resulted from the felling of primary forest 
for native gardens; the frogs w^ere also found in cavities in the rotting 
logs which were lying about. 

Arthroleptis moorii Boulenger 

Arthroleptis moorii Boulenger, 1898, Proc. Zool. Soc. London, p. 479, pi. 
xxxviii, fig. 2: Kinyamkolo, Lake Tanganyika. 

14 (M. C. Z. 17127-36) Nyamkolo, Lake Tanganyika. 9. v. 30. 

Distribution. This little frog is, I believe, still only known from the 
type of which the present series are topotypes for the prefix "Ki" in 
the local dialect is an augmentative, in direct contrast to its use as a 
diminutive in Swahili. Niomkolo is another rendering of the name of 
this same village which is situated on the southeast shore of Lake 
Tanganyika in Northern Rhodesia. 

Variatio7i. I should consider the nostril equidistant from the eye 
and end of snout ; while the first and second finger are sometimes of 
equal length, in some specimens the first is very definitely shorter 
than the second ; the tips of the digits are not swollen ; the tibio-tarsal 
articulation, while generally only reaching the eye or nostril, does some- 
times reach the end of the snout as in the type. 

Coloration in life. The pattern is exactly like that of the holotype 
in some specimens, in others there is a narrow vertebral stripe. Above, 
a pale ashy-grey or grey-brown with blotches and bars in sepia; the 
interorbital mark mentioned by Boulenger is more four-sided in the 



384 bulletin: museum of comparative zoology 

present series, both the elongate lateral and the shorter posterior 
edges being concave in outline and the four corners produced into 
points and heavily pigmented; lips speckled with white. Below, white, 
the throat flecked with dusky, this pigmentation may be so concen- 
trated in males as to give the impression of a grey throat. 

Measurements. The largest is a male measuring 17 mm., there are 
no adult females. 

Breeding. The high propoi:tion of \er\ young frogs is evidence that 
the breeding season is past; the rains had ended in March. 

Habitat. By reason of their small size these frogs are exceedingly 
difficult to find. The whole series was taken in the dew-laden short 
grass about the centre of the bay immediately below the residential 
buildings of the mission. If one of these frogs is disturbed it takes but 
one hop then dives in among the roots of the grass where it remains 
motionless. The collecting of this series represents several hours of 
concentrated search whereas if Xyamkolo had only been visited at 
the beginning of the breeding season when presumably moorii would 
be congregating in pools it is probable that an adequate series might 
have been obtained with ease. 

Arthroleptis minutus Boulenger 

Arthroleptis minutus Boulenger, 1895, Proc. Zool. Soc. London, p. 539: Durro, 
western Somaliland; Loveridge, 1929, U. S. Nat. Mus. Bull. No. 151, p. 
107: Localities in L'ganda, Kenya and Rhodesia. 

9 (M. C. Z. 17051-9) Bagamoyo. 12. xi. 29. 
80 (M. C. Z. 17060-3) Dabaga, Uzungwe Mtns. 1. i. 30. 
49 (M. C. Z. 17064-74) Kigogo, Uzungwe Mtns. 13. i. 30. 

1 (M. C. Z. 17075) Mwaya, Lake Nyasa. 1-8, iii. 30. 

4 (M. C. Z. 17080-3) Kipili, Lake Tanganyika. 19. v. 30. 

5 (M. C. Z. 17084-8) Albertville, Lake Tanganyika. 21. v. 30. 

2 (M. C. Z. 17089-90) Entebbe, Lake Victoria. 27. vi. 30. 

Native name. Kasambara (Kinyakusa). 

Coloration in life. When at Dabaga, I had not realised that the two 
types referred to in the following note were two species, the larger 
minutus, the smaller parvulus. The note reads: — "There seem to be 
two types of males, a larger, which is lemon-yellow on the belly, bright 
yellow on the throat, and a smaller which has a white, or bluish-white 
belly mottled with black and a black throat, is it possible that the 
latter are immature for their throats only show a bagginess while the 
yellow-throated males have a definite disk." Unfortunately in alcohol 
the whole Dabaga series of males are uniformly white below. 



loveridge: African herpetology 385 

On reaching Kigogo 1 found the two species occupying separate 
breeding areas and so recognized the two species. The color of Kigogo 
frogs of the species minutus was: cf. Throat bright chrome, lemon- 
yellow or whitish on the belly with very few specklings and such as 
there are, are lateral. 9 . Below, white or silvery -w^hite, specklings 
when present are also lateral as in the males. 

Measurements. The largest male is IS mm.; the largest female 22 
mm. ; tlie smallest frogs are 9 mm. 

Breeding. Undoubtedly assembling to breed at both Dabaga and 
Kigogo. At the former locality the series consisted of 64 males and 
16 females, at the latter 20 males and 28 females; these females, as is 
also one from Entebbe, are distended with ova except a few at Kigogo 
which appeared to have laid. 

The Kipili and Albertville frogs are very young and identified with 
doubt as they may w^ell be referable to parvulus; those from Kipili 
measure 13 to 14 mm. in length while the range of the Albertville 
series is 9 to 11 mm. 

Call note. The call of minutus is \ery like the rattling call of Phry- 
nohatrachus acridoides, but ends in a click; it is, of course, much fainter. 

Parasites. The chigger mites which I mentioned in 1929 (loc. cit.) 
have since been described as Endutromhicida penetrans by Dr. H. E. 
Ewing. 

Enemies. At Mwaya, one was taken from an Olive Water Snake 
{Natrix olivaceus) and another from a Hissing Sand Snake {Psammo- 
phis sibihms). 

Habitat. The Bagamoyo frogs were found in a sw^amped mbugwe 
three miles out along the Dar es Salaam road and by waterholes near 
the seashore. This frog does not like the water but may be found in 
damp grass or on mud in the vicinity of w^ater holes. Those from 
Dabaga were taken in cattle-trampled bog land bordering a swiftly- 
flowing brook in the bottom of a valley. 

Arthroleptis ukingensis Loveridge 

Arthrolepiis ukingensis Loveridge, 1932, Bull. Mus. Comp. Zool., 72, p. 385: 
Madehani, Ukinga Mtns., southwestern Tanganyika Territory. 

3 (M. C. Z. 17137-9) Madehani, Ukinga Mtns. 14. ii. 30. 
3 (M. C. Z. 17076-8) Ilolo, Rungwe district. 15. iii. 30. 

Remarks. As stated in the diagnosis, this frog is near A minutus 
Boulenger from which it is distinguished by very well developed digital 
expansions. In Phrynobatrachus acridoides there is wide variation in 



386 bulletin: museum of comparative zoology 

this character but there are no expansions in any of the minutus which 
I have collected in East Africa over a long period. 

It agrees closely with De Witte's description of A. houlengeri from 
the St. Louis Plain at the edge of Lake Tanganyika excepting that 
houlengeri lacks an outer metatarsal tubercle as well as a tarsal tubercle. 

Ahl has described a related form from Buala in the Cameroons as 
A. j^i/gmaeus but the Madehani frogs differ from his description of the 
14 mm. holotype in possessing a lingual papilla; in the snout being 
longer than the eye diameter; the 3rd finger being as long as the snout; 
the tibio-tarsal articulation reaching the eye and minor differences so 
that it is improbable that they are specifically identical. 

IVIr. H. W. Parker has examined the Madehani series which he 
considers conspecific with the frogs from Kibonoto referred by Lonn- 
berg to "^4. bottegt'' which he feels sure is an incorrect indentification. 
He adds that after comparing the Kibonoto frogs and 17137-9 with 
the type of minutus " I can only find a most trivial difference in the 
tarsal tubercle. In the type it is more a triangular flap-like, continua- 
tion of the tarsal fold, whereas in the others it is a conical papilla." 

I quite agree that little importance can be attached to the appear- 
ance of the tarsal tubercle for wide variation occurs in its development 
in our large series of minutus. On geographical grounds I should 
imagine the Kibonoto frogs to be referable to minutus. 

Breeding. The type was taken in embrace with a male in a small 
patch of swamped ground close to some wild bananas growing on the 
mountain side without the forest. Beside them was some jelly, con- 
taining tadpoles, which was plastered over leaves and twigs ; it would 
have been almost impossible to find clear water at this spot. 

Parasites. Ilolo frogs have minute dermal parasites on the skin of 
the belly. 

Arthroleptis rungwensis Loveridge 

Arthroleptis rungwensis Loveridge, 1932, Bull. Mus. Comp. Zool., 72, p. 386: 
Ilolo, Rungwe Mtn., southwestern Tanganyika Territory. 

9 Type (M. C. Z. 17141) Ilolo, Rungwe Mtn. 15. iii. 30. 

The complete discussion about this frog appears in the citation given 
above. 

Arthroleptis parvulus Boulenger 

Arthroleptis parvulus Boulenger, 1905, Ann. Mag. Nat. Hist. (7), 16, p. 109, pi. 
iv, figs. 3-3b: Bange Ngola northeast Angola. 



loveridge: africax herpetology 387 

2 (M. C. Z. 17091-2) Dabaga, Uzungwe Mtns. 1. i. 30. 
70 (M. C. Z. 17093-100) Kigogo, Uzungwe Mtns. 24. i. 30. 

5 (M. C. Z. 17101-5) Lukungu, Ubena Mtns. 8. ii. 30. 

1 (M. C. Z. 17106) Mangoto, Ukinga Mtns. 10. ii. 30. 
11 (M. C. Z. 17107-16) Tandala, Ukinga Mtns. 11. ii. 30. 

1 (M. C. Z. 17117) Bulongvva, Ukinga Mtns. 12. ii. 30. 
Tadpoles & 8 (M. C. Z. 17118-26) Madehani, Ukinga Mtns. 14-19. ii. 30. 

Distribution. This species is new to Tanganyika Territory being 
known only from northeast Angola and the southern Belgian Congo. 

Native name. Bungulula (Kikinga). 

Affinities. So closely related is this species to A. minutus that I can 
find no characters, other than size and breeding coloration and call 
notes, to distinguish them. As already related, both species are to be 
found in the same swamp at Dabaga, but at Kigogo they occupied 
different valleys and the difference in the calls was very plain. It may 
be that some parvulus are still included in the big series of minutus 
from Dabaga, but the Kigogo specimens were sorted in the field and 
retain their characteristic ventral pigmentation. 

Variation. They agree closely with the description of the types. 

Measurements. The largest of sixty males in the Kigogo series 
measured 15 mm., the largest of the ten females in the same series 
measure 18 mm., these were measured in the field. The rest of the 
series consists of fifteen males ranging from 13 to 16 mm., with an 
average of 14 mm., seven females ranging from 13 to 17 mm. with an 
average of 15 mm. and two young, 8 and 10 mm. in length. 

Breeding. Three pairs were taken in embrace at Kigogo where the 
swamps resounded with their vibrant trilling calls and the males out- 
numbered the females by six to one. Three weeks later a pair was 
taken at Madehani together with a young one of 8 mm. The tadpoles 
taken at the same time may be those of schubotzi which was found in 
the same swamp. 

Habitat. Most of the Kigogo series were found in a peaty drainage 
area where there were a few small catchments of water. The surround- 
ing ground was so damp as to well up with moisture at every footstep. 
Such a situation was wholly typical of their haunts in the ravines and 
shallow valleys of these uplands. Five were taken at Lukungu in as 
many different swamps but A. minutus was neither heard nor seen. 
I noted that parvulus was present in all the swamps and streams along 
the trail to Mangoto. 



388 bulletin: museum of comparative zoology 

Hemisus marmoratum marmoratum (Peters) 

Engystoma marmoratum Peters, 1855, Arch. Naturg., 21, part 1, p. 58; Caba- 

geira, Mozambique. 
Hemisus marmoratum Nieden, (part), 1926, Das Tierrich, Anura 2, 49, p. 11: 

Wager, 1929, Trans. Roy. Soc. S. Africa, 17, pp. 127-135, text figs. 1-5, 

plates viii-x. 

Eggs, tadpoles and 11 (M. C. Z. 16455-67) Bagamoyo. 13. xi. 29. 

Distribution. Nieden has recorded this race from Mwanza. In addi- 
tion to the Bagamoyo specimens listed above, I have utiHzed eleven 
other examples from Tanganyika Territory which are in the collection 
of the Museum of Comparative Zoology. They are from Gonya, 
Tanga, Dar es Salaam, Morogoro, Kilosa, Uliya, and Sagayo in 
Mwanza district. 

Affinities. De Villiers considers Hemisus a ranid, rather than a 
brevicipitid, but Noble prefers to continue its inclusion in the Brevi- 
cipitidae. Parker, however, agrees with De Villiers. The reason for 
employing trinomials is explained under H . m. guinecnsis Cope. 

Measurements. Nine males measure from 24 to 27 mm. 
Nine females " " 27 to 33 mm. 

Four young " "17 to 23 mm. 

The snout is contained in the total length from 6 to 8 times, 
tibui 1.1 to l.i times. 

" i?„„x ii ii II a i( it << 1 ,1 X 1 ■" x' 

toot 1.4 to 1./ tmies. 

Breeding. Bagamoyo is so dry that the natives set their banana 
plants in hollows when possible. Digging at the roots of these bananas 
we discovered a female with her eggs just two inches below the surface 
of the sand, six inches away was a male. At the base of a second 
banana another female was discovered resting on a mass of jelly or 
slime in which tadpoles were actively moving. The mucous or gela- 
tinous substance was intermixed with sand and though I carefully 
transferred the whole mass to a tin, by the time camp was reached 
nothing but tadpoles and sand remained. The tadpoles were picked 
out and a hundred and ten recovered, evidently Hemisus deposits a 
large number of eggs at one time. The nearest male found by the spot 
where these tadpoles were discovered was six feet away. The interest- 
ing breeding habit of Hemisus has long been known. A s an inhabitant 
of sandy, and sometimes desert regions, the advantages to it of a 
method which dispenses with surface water during the early stages of 
development are obvious. There was no water in the hollow at this 
time and, owing to the sandy terrain, it is doubtful if water would be 



loveridge: African herpetology 389 

retained for more than a few hours after a rainstorm. For a detailed 
account of the interesting nesting habits of this creature, see Wager's 
paper cited above. 

Did. Termites in those examined. 

Hemisus marmoratum guineensis Cope 

Hemisus g^nneeyisis Cope, 1865, Nat. Hist. Review, p. 100, footnote: presum- 
ably Guinea (Type in the Vienna Museum). 

Hemisus marmoratum Noble, part, 1924, Bull. Am. Mus. Nat. Hist., 49, p. 279: 
Niangara, Faradje and Zambi, Belgian Congo. 

4 (M. C. Z. 16468-71) Masiliwa, Turn. 9. xii. 29. 
1 (M. C. Z. 16472) Mangasini, Usandawi. 14. xii. 29. 

J^ariatiott. Coming fresh from collecting a series of breeding females 
and eggs of //. m. marmoratum at Bagamoyo where none exceeded 
33 mm. in length from snout to anus, I was astonished to find at 
MasiHwa females measuring 32, 44, 45 and 49 mm. respectively. 

Consulting Noble's most useful report on the variation which he 
observed in ninety-six specimens from the Congo, I observe that he 
states, " Gravid females average 47 mm. (maximum 50 mm. ; minimum 
45 mm.)." This caused me to measure the four West African (Belgian 
Congo, Nigeria and Liberia) specimens in our collection and found that 
the West African frogs are much larger than typical marmoratum from 
the east coast, and may be separated by the number of times the length 
of the foot is contained in the length from snout to anus. Parker in- 
forms me that of a pair from Pako, Ituri, Belgian Congo, the cf 
measures 35 mm., the female 52 mm. 

In looking for the name which should be employed, I cannot use 
H. sudanensc (Steindachner) for not only is it of the same dimensions 
as marmoratum but our single example from Kordofan bears out the 
author's remarks and I think that it may be safely considered as a 
subspecies, viz. //. m.. sudanense (Steindachner) distinguished from the 
typical form by its longer and more shovel-shaped snout ; it certainly 
(liffers in these characters from the thirty-one members of the genus 
in the Museum of Comparative Zoology. 

E. guineensc, a m.s.s. name attributed to Sir A. Smith by Giinther in 
1858 (Cat. Batr. Sal. Brit. Mus., pp. 47 and 137) is devoid of descrip- 
tion but Cope's footnote (1865) validates the name for he refers to 
certain anatomical characters of a specimen in the Vienna Museum. 
Cope therefore becomes the author. The two forms may be differ- 
entiated thus: 



390 bulletin: museum of comparative zoology 

Foot contained in length from snout to anus 1.4 to 
1.7 times; adult males 24 to 27 mm., adult 

females 27 to 33 mm m. marmoratus 

Foot contained in length from snout to anus 1.7 to 
1.9 times; adult males 31 to 36 mm., adult 

females 32 to 52 mm m. gmneemis 

Measurements. One male measures 36 mm. (Noble's series 31 to 
34 mm.) Six females measure 32 to 49 mm. (Noble's series 45 to 50 mm.) 
The snout is contained in the total length from 7.3 to 8.8 times. 
" tibia " " " " " " " 2.2 to 2.8 times. 

" foot " " " " " " " 1.7 to 1.9 times. 

Coloratiofiinlife. Handsomelyspotted and streaked with pale yellow. 
Habitat. The Masiliwa specimens w^ere taken during a shower at 
9 p.m. as they were hopping about on recently hoed ground as de- 
scribed in detail under Rana delalandii. 



POLYPEDATIDAE 

Chiromantis petersii petersii Boulenger 

Chiromantis petersii Boulenger, 1882, Cat. Batr. Sal. Brit. Mus., p. 93, pi. x, 
figs. 1-la: "Interior of East Africa," i.e. Mpwapwa, Ugogo, Tanganyika 
Territory; Loveridge, 1929, U. S. Nat. Mus. Bull. 151, p. 119: Localities 
in Tanganyika Territory. 

Chiromantis pygmaeus Ahl, 1930, Zool. Anz., Berlin, 88, p. 219: Kibwezi, Kenya 
Colony. 

Chiromantis pidus Ahl, 1931, Sitz. Ges. Naturf. Freunde Berlin, p. 213: Kili- 
matinde, Tanganyika Territory. 

Chiromantis rugosus Ahl, 1931, Sitz. Ges. Naturf. Freunde BerUn, p. 215: 
Langenburg (i.e. Manda), Lake Nyasa, Tanganyika Territory. 

1 (M. C. Z. 16735) KUimatinde, Ugogo. 27. xi. 29. 
1 (M. C. Z. 16736) Unyanganyi, Turu. 6. xii. 29. 
10 (M. C. Z. 16737-46) Handa, Usandawi. 10. xii. 29. 
Nest, eggs, tadpoles & 32 (M. C. Z. 16747-59) Mangasini, Usandawi. 13-14. 

xii. 29. 
7 (M. C. Z. 16760-6) Shinyanga, Usukuma. 3. vi. 30. 

Variation. The Mangasini series alone cover all the variations for 
which Ahl proposed the names albescens and fasdatus in 1929 and con- 
firm the view that all these alleged " species " are synonyms of petersii. 
The tibia is included from twice (M. C. Z. 16750. d' ) to two and a half 
times (M. C. Z. 16754. 9 ) in the length from snout to anus. I was, 



loveridge: African herpetology 391 

however, in error in synonymising macrops with typical petersii; it 
should be referred to the synonymy of C. petersii kcllcri Boettger. 

More recently Dr. Ahl has proposed three more names based on 
trivial variations which it is not necessary to discuss. It will be 
observed that jNI. C. Z. 16735 is a topotype of pidus and it might be 
pointed out that Kilimatinde is only a hundred miles from the type 
locality of petersii, that the official altitudes are 3,591 feet for Kili- 
matinde and 3,312 feet for Mpwapwa and that both are in the same 
continuous stretch of thorn-bush steppe with apparently identical 
ecological conditions. On topographical grounds more material may 
demonstrate rugosus to be a southern form of petersii. 

Coloration. The usual hue of this frog may be said to be clay- 
colored. Such a frog found squatting in the direct rays of the sun at 
9.30 a.m. was placed in a white bag where it remained for twenty -four 
hours; on removal it was found to be a wood-brown shade. Breeding 
males are distinguishable by their throats being plumbeous colored. 

Measurements. Omitting the Shinyanga series which range from 35 
to 42 mm., twenty males range from 41 to 49 mm., with an average of 
45 mm. ; twenty-seven females range from 46 to 65 mm. with an average 
of 56 mm . 

Breeding. The Kilimatinde and Unyanganyi females are full of ova. 
At Handa nine frogs were found squatting on thorn-bush twigs around 
a water hole and one on the bare ground near a shallow pool. It was 
interesting to note that though there were shady, leafy posts around 
these water holes the frogs preferred to squat on the bare, dead, thorn 
bush employed to reinforce the stockade and which was in the full 
glare of a sun that seemed exceptionally hot. 

Eight nests were found around these w ater holes, either in the fresh 
green grass of which there were one or two patches, or in niches of the 
rough earth banks, or simply on mud at the water's edge. Five nests 
were seen on the mud round a shallow rain pool out on the mbugwe 
(black cotton soil plains with scattered bull's-horn acacia) ; these were 
all more or less trampled into the mud and destroyed by game coming 
down to drink. Seventeen nests were counted on the mud surrounding 
one large pool in the bottom of a ravine that would be the bed of a great 
river during the rainv season. In none of these sites were there bushes 
within five feet of the water's edge, the eggs were yellow spheres and 
evidently very recently deposited. 

At Mangasini, on December 13th, I found twenty-five of these frogs 
on thorn bush and scrubs in the vicinity of a water hole. The females 
were distended with ova. Thirty feet from the water none were to be 



392 bulletin: museum of comparative zoology 

found. Presumably they had congregated as a result of the eighteen 
hours continuous downpour with which the rainy season started yester- 
day, the shower only terminating at noon today (13th). At 8 p.m. I 
went down in bright moonlight but could find no frogs in the water and 
only a little desultory calling was proceeding from the thorn bush fence 
surrounding the water hole. The call is a " wock-wock" sometimes pro- 
longed with a whirr. During the night these frogs spawned on the mud 
surrounding a pan of water which I had put in a temporary vivarium. 

Next morning (14th) I took seven frogs in a large shallow pool in 
the swamped mhugwe and we found two nests in tussocks of grass 
growing from the water. On one of these nests a female still sat, vigor- 
ously but mechanically working up the foam with her hind legs, her 
left hand rested in a fork of the reedy grass and her right hand grasped 
some grass with the fingers all in one plane. She was six inches above 
the surface of the water, this measurement corresponded to the height 
of the nest. It was observed that fresh foam is pure white and assumes 
its creamy tint only on drying, it is very glutinous when fresh and this 
stickiness affords an excellent protection from insects, for one or two 
were seen trying to free themselves ha\'ing inadvertently alighted upon 
it. I have seen beetles run freely over a dry nest. 

On December 17th, when twenty miles north of Saranda, I found 
quite fifty nests around a very large pool, all laid on the mud though 
branches were available. Some nests appeared to be the work of se\^eral 
frogs for they would easily have filled an ordinary washing basin. Two 
frogs were seen but not collected. On the 18th, at Saranda, several 
nests were found on the trampled edges of a pool. It is difficult to see 
much survival value in the unintelligent method of depositing eggs as 
practiced by Peter's Frog in the semi-arid thorn-bush country ; great 
numbers of nests are destroyed by the hooves of game or cattle and 
myriads of tadpoles perish by stranding on the mud. 

Aestivating ? The countryside was already becoming parched on 
June 3rd when we discovered five frogs among the twigs forming the 
basal part of a crow's nest at Shinyanga. The nest was situated thirty 
feet from the ground in the bare branches of a baobab tree growing in 
open country. These toads were all of small size being 35 to 39 mm. 
in length. Dissection of two showed their stomachs to be empty and 
quantities of bright yellow fat were present, though lacking in the 
breeding frogs taken elsewhere. The rains had ceased a month before, 
the weather was already very hot, and it seemed probable that the 
frogs were aestivating. 

Ilabitaf. A search of IVIanyara hedges at Dodoma in mid-November 



loveridge: African herpetolouy 393 

failed to discover any of these frogs. At Kilimatinde I examined nearly 
a mile of hedge before securing one. 

Leptopelis bocagii (Giinther) 

Cystignathus bocagii Giinther, 1864, Proc. Zool. Soc. London, p. 481, pi. xxxiii, 

fig. 2: Duque de Bragan^'a, Angola. 
Hylambates margitiatus Bocage, 1895, Herpetologie D' Angola, p. 178: Quissange, 

interior of Benguella, Angola. 
Leptopelis bocagii Loveridge, 1929, U. S. Nat. Mas. Bull. 151, p. 121: Ulukenya 

Hills, Kenya Colony. 
Hylambates brevipalmatus Ahl, 1930, Zool. Anz., 87, p. 228: Unyika, Tanganyika 

Territory. 

2 (M. C. Z. 16767-8) Masiliwa, Turn. 19. xii. 29. 

1 (M. C. Z. 16769) Senjeri Pass, s. of Unyika. 5. v. 30. 

Affinities. The specimen from Senjeri Pass is almost topotypic of 
Ahl's H. brevipalmatus from the description of which it differs in the 
hind foot being included L64 times in the length from snout to anus 
which is 2.5 times in hrnnpalmatus according to Ahl. A request for in- 
formation on this point was not favored with a reply. ]VIore recently, 
at my suggestion, Mr. K. P. Schmidt kindly reexamined the type and 
remeasured it so that I find that the length of the foot is really included 
1.61 times in the body length. Mr. Schmidt's measurements of the type 
of breripalmafiis are : Length from tip of toe to heel 25.5 mm. Length of 
longest toe from its tip to the posterior edge of the large metatarsal 
tubercle 18 mm. Finally Mr. Schmidt states that the shoulder girdle of 
the type, though somewhat mangled, is that of Leptopelis and not of 
Hylambates. He concurs in synonymising it with bocagii. 

Boulenger (1906, Ann. Mus. Civ. Stor. Nat. Gen. (3), 2, (42), p. 10) 
remarks that H. marginatus Bocage, which was only known to him 
from the description, "appears to differ in the longer toes (foot more 
than half the length of head and body) ; if one works out the propor- 
tions given by Bocage (1. c. pp. 176-180) it will be seen that this 
assumption was incorrect for H. marginatus is 1.80 times in the length 
and the other species already referred to the synonymy of bocagii by 
Boulenger after " careful comparison" of the types and " a large series 
of specimens" are H. bocagii, 1.53 times, H. cinnamomeus Bocage, 1.44 
times, //. anchietae Bocage, 1.96 times. Thus the feet of the types of 
these Angolan specimens range from 1.44 to 1.96 times in the length 
from snout to anus. As a check the series of eight .specimens in the 
Museum of Comparative Zoology ha\e been measured and found to 



394 bulletin: museum of comparative zoology 

range from 1.43 times (Faradje, Belgian Congo) to 1.75 times (Guaso 
Nyiro, Kenya Colony); the Tanganyika series alone from 1.52 to 1.64 
times. 

Variation. The tibio-tarsal articulation falls short of the tympanum 
but the metatarsal articulation marks the eye; the interorbital space 
equals the width of an upper eyelid. 

Coloration. All possess the characteristic dark dorsal patch, but 
posteriorly the patch tends to break up. 

Measurements. These three males range from 45 to 52 mm. 

Diet. The Masiliwa frogs taken at 8 a.m., held fourteen long termites, 
a beetle, earwig, spider and centipede. The Senjeri frog, on the other 
hand, taken by a roadside ditch at 11 p.m., had an empty stomach 
but was loaded with fat wdiich was lacking in the Masiliwa frogs. 

Enemies. The Masiliwa frogs were taken from the mouth and stom- 
ach of a Boomslang {BisphoUdus ti/pus) which I captured in the act of 
swallowing one of them. 

Leptopelis johnstoni (Boulenger) 

Hylmnhates johnstoni Boulenger, 1897, Proc. Zool. Soc. London, p. 803, pi. xlvi: 
Kondowe to Karonga & Nyika Plateau, Nyasaland. Lcinnberg, 1907, 
Reptilia and Batrachia in Sjostedt, 1910 Kilimandjaro-Meru Exped., 1, 
part 4, p. 25: Mombo, Usambara, Tanganyika Territory. 

Leptopelis johnstoni Barbour & Loveridge, 1928, Mem. Mus. Comp. Zool., 50, 
p. 239: Bagilo, Uluguru Mtns., Tanganyika Territory. 

2 (M. C. Z. 16770-1) Mwaya, Lake Nyasa. 1. iii. 30. 
1 (M. C. Z. 16772) Ilolo, Rungwe district. 8. iv. 30. 

Distribution. ]Mwaya is only a few miles north of Karonga and Ilolo 
is practically on the Xyika Plateau so that these specimens are almost 
topotypes. 

Affinities. These frogs agree with vermiculatus from the nearby 
forest in their strongly compressed tubercle and in the interorbital 
width being the same as an upper eyelid but differ in vomerine teeth, 
smaller disks, less extensive webbing of the hind feet and coloration. 

Variation. The tibiotarsal articulation reaches the eye in the 
Mwaya frogs (cf , 9 ) but falls short in the Ilolo specimen ( 9 ). 

Coloration in life. Agrees with that described for the Uluguru 
specimens. 

Measurements. The male is 42 mm., the large female 62 mm. 

Breeding. The Mwaya female holds a mass of eggs, each about 2.5 
mm. in diameter: the Ilolo frog seems to have spawned quite recently. 



LOVERIDGE: AFRICAN HERPETOLOGY 395 

Diet. A very large snail was taken from the Ilolo frog's stomach. 
Habitat. All were taken in domestic bananas. 

Leptopelis vermiculatus (Boulenger) 

Hylambates vermiculatus Boulenger, 1909, Ann. Mag. Nat. Hist. (8), 4, p. 497: 

Amani, Usambara Mountains, Tanganyika Territory. 
Leptopelis signifer AM, 1929, Sitzber. Ges. Natui-f. Freunde Berlin, p. 216: 

Amani, Derema, etc., Usambara Mountains, Tanganyika Territory. 

2 (M. C. Z. 16773-4) Nkuka Forest, Rungwe Mtn. 9. iv. 30. 

Distribution. Nieden referred two young frogs from Uhehe to this 
species which is very closely related to rufus. 

Affinities. In an attempted revision of the genus Leptopelis, of 
which less than a dozen species were known in 1928, Dr. Ahl has 
described no fewer than twenty-one additional species. It is obvious 
that Dr. Ahl's conception of what constitutes a species differs widely 
from that of his predecessor Dr. F. Nieden, for the types of fourteen of 
these "species" had been identified previously with rufus by Nieden. 
It is also interesting to note that Amani or the Usambara Mountains 
is type locality for six of these species. 

The key which is supposed to enable one to distinguish these species 
is almost entirely useless in practice, being based on the most trivial 
differences long known to vary with age and sex. It is perfectly easy 
to take a series of frogs from one locality like Bagilo in the Uluguru 
Mountains and on applying the key find that they break up into several 
species or land in deadlocks. It is not surprising therefore to find that 
eleven of these new "species" were based on single frogs. 

When describing L. signifer Dr. Ahl designates three Derema speci- 
mens as "Type" (presumably meaning cotypes) and regards fifteen 
others as "Cotypes" (i.e. paratypes). Derema is scarcely two miles 
from Amani. A Derema cotype of signifer (M. C. Z. 17530) has been 
compared with an Amani topotype of vermiculatus (M. C. Z. 13598) 
which in its turn had been compared with the type in the British Mu- 
seum. There seems to be no reason for supposing that signifer is 
anything but a strict synonym and one wonders how it came to be 
described. 

Coloration in life. Above, purplish-brown but the whole of the 
centre of the back is occupied by an arrow-head marking as in L. john- 
stoni, the apex reaching nearly to the occiput; this marking is a rich 
olive green bordered by black; there are three patches of the same color 
on the lips, the last prolonged posteriorly over the tympanum nearly to 



396 bulletin: museum of comparative zoology 

midbod y ; two large, rather brighter blotches further Ijack on the flank, 
the hindmost extending upwards to coalesce with the arrow-head 
marking above the groin; all these markings are edged with black; the 
fore and hind limbs are green, barred with darker; there are five cream 
colored spots on the upper lip of which the middle one extends up- 
wards to merge into the brown of the head ; the disks of the two inner- 
most fingers are cream colored. Below, cream and pure white marbled 
with purplish and greenish-brown, the abdomen tinged with ochre; a 
rather ill-defined cream -colored streak along the underside of the hind 
limb and outer edge of the foot; no markings on anus or heel. 

Measuremenis. The adult female measures 65 mm. , immature female 
55 mm. 

Breeding. The ovaries of the larger frog hold developing ova. 

Parasites. There are minute, pink, larval mites embedded in the 
skin of the feet. 

Habitat. When forced by a downpour to take refuge in a saw-pit 
which was roofed over, I discovered the female adult snugly ensconced 
in a little depression beneath the drift of dead lea\es which covered 
part of the floor. The leaves were mostly brown or black so that the 
brilliant coloring of the frog was not a little astonishing. 



Leptopelis aubryi (A. Dumeril) 

Hyla aubryi A. Dumeril, 1856, Rev. Mag. Zool. (2), 8, p. 561: Gaboon. 
Leptopelis tessmanni Noble {nee. Nieden), 1924, Bull. Am. Mus. Nat. Hist., 

49, p. 245: Medje, Belgian Congo. 
Leptopelis aubryi Barbour & Loveridge, 1928, Mem. Mus. Comp. Zool., 50, p. 

233: Mt. Lutindi, Usambara Mtns., Tanganyika Territory. 
Leptopelis barbouri Ahl, 1929, Sitzber. Ges. Naturf. Freunde Berlin, p. 199: 

Mt. Lutindi, Usambara Mtns., Tanganyika Territory. 

Affinities. Two of the series of Medje frogs referred to tessmanni by 
Dr. G. K. Noble are now in the Museum of Comparative Zoology and 
I consider are indistinguishable from our series of Cameroon L. aubryi. 

In 1928, when Dr. T. Barbour and I referred certain frogs from Mt. 
Lutindi to L. aubryi we drew attention to the fact that the tympana 
were lacking or concealed in the nine young while those of the adults, 
or semi-adults, which were taken within fifty feet of the young, were 
three-quarters and seven-eighths the diameter of the eye, instead of 
"half" as stated by Boulenger in 1882. 

It was supposed that it would be apparent to most herpetologists 
that if the tnnpanum is concealed in the young, its proportions, at 



loveridge: African herpetology 397 

various stages of development, in relation to the eye diameter must be 
radically different. This rather obvious inference apparently escaped 
Dr. Ahl who, without ever having examined them and on the basis of 
the four lines of comment which we made, designates these specimens 
as types of a new species which he calls Leptopelis harhouri. 

I might add that Dr. Barbour is in entire agreement with m^- action 
in synonymising this name; we have had the somewhat unique advan- 
tage over the author in having seen the types and after careful com- 
parison with a series of L. auhryi from the Cameroons, we fail to detect 
any structural differences which would warrant the assumption that 
we are dealing with two distinct species or recognisable geographical 
races. 

Leptopelis uluguruensis Barbour & Loveridge 

Leptopelis uluguruensis Barbour & Loveridge, 1928, Mem. Mas. Comp. Zool., 
50, p. 235, pi. iii, fig. 3: Nyange, Uluguru Mountains, Tanganyika Terri- 
torj'. 

Leptopelis tanganus Ahl, 1929, Sitzber. Ges. Natui'f. Freunde Berlin, p. 221: 
Amani, Buloa (i.e. Bulwa), and Tanga, Tanganyika Territory. 

Through the courtesy of Dr. Ernst Ahl a cotype of his L. tangambs 
has been received and carefully compared with the type series of L. 
uluguruensis from which it does not appear to differ. 



Megalixalus fornasinii (Bianconi) 

Euchnemis fornasinii Bianconi, 1850, Spec. Zool. Mosamb. Rept. pi. v, fig. 1 : 

Mozambique. 
Megalixalus fornasinii var. unicolor Boettger, 1913, in Voeltzkow, Reise in 

Ostafrika, p. 349: Pemba Island. 
Megalixalus loveridgii Procter, 1920, Proc. Zool. Soc. London, p. 418: Morogoro 

Tanganyika Territory; Barbour & Loveridge, 1928, Mem. Mas. Comp. 

Zool., 50, pp. 227-230: Many localities. 
Megalixalus fornasinii Parker, 1930 (1931), Proc. Zool. Soc. London, pp. 900- 

902: Localities in Mozambique. 
Megalixalus dorsimaculatus Ahl, 1930, Sitzber. Ges. Naturf. Freunde Berlin, 

p. 92: Magrotto near Tanga, Tanganyika Territory. 
Hyperolius pygmaeus Ahl, 1931, Mitt. Zool. Mus. Berlin, 17, p. 22: Tanga, 

Tanganyika Territory. 

4 (M. C. Z. 16801-4) Bagamoyo. 16. xi. 29. 
34 (M. C. Z. 16805-17) Morogoro, Ukami. 20. xi. 29. 
7 (M. C. Z. 16818-25) Mwaya, Lake Nyasa. 1-8. iii. 30. 



398 bulletin: museum of comparative zoology 

Distrihution. Recorded from Bagamoyo, Ukami, Ukinga ^Mountains 
and Rungwe by Nieden. 

Native name. Pasa (Kinyakusa). 

Affinities.. In 1930 Ahl attempted a revision of the genus Megalixalus 
with results that can only be characterized as deplorable for he recog- 
nized almost every species described, reviving M. spinifrons (Cope) 
and M. stuhlmatuii Pfeffer which have long been recognized as strictly 
synon.Ainous svith fornasinii. Xo notice is taken of the transference of 
Megalixalus gramineus to Leptojjelis by Parker, " Hyperolius" fulvo- 
vittatus Cope is omitted, etc. etc. Hyperolius pygmaeus appears to be 
based on a young male, the only specimen was 17 mm. An 18 mm. male 
(M. C. Z. 16S05) in the present series agrees so closely structurally, disk 
and all, with Ahl's description as to make me certain they are syn- 
onymous. The white flecked, broad, brown lateral band is characteris- 
tic of Tanga specimens. 

More recently, January 1931, Parker cleared up the tangle in which 
certain species of the genus have been since 1882 by showing that 
M. dorsalis (Peters) of West Africa is a valid species (it was recognized 
by Ahl as distinct as he considered nearly all names valid) long con- 
fused with M. foniasinii of Avhich M. loveridgii is a straight synonym. 

Variation. The spines of three Bagamoyo frogs were flush with the 
surface of the skin and had only the appearance of minute dots. 

Coloration. Only one of the four Bagamoyo frogs had a mid-dorsal 
streak, the others were silvery -white soon changing to brown when 
put in a bag. The jNIorogoro series, topotypes of loveridgii, show the 
same variation, some having a dorsal streak like the type, others being 
uniform like our cotype of Boettger's unicolor from Pemba which 
that author recorded as occurring \\\th fornasinii on the island. 

I might add that when sending me this cotype of imicolor in 1929, 
Dr. Robert ]\Iertens told me that he considered /or/^o^i/in, unicolor and 
loveridgii were all one species and that he imagined that the minute 
spines probably became more prominent during the breeding season. 
He was entirely correct. 

Breeding. The presence of so many 15 mm. young at Morogoro, as 
descriljed below, indicates that the breeding season was recently over 
in that locality, i.e. had taken place during the " big rains." 

Enemies. At Mwaya a frog was recovered from the stomach of a 
Spotted Wood Snake (Philothamnus s. dorsalis). 

Habitat. I took the first pair of Bagamoyo frogs in the central shoot 
of a domestic banana, a third was in a similar situation but the fourth 
came from sedges at the edge of a swamp. The species appeared dis- 



loveridge: africax herpetology 399 

tinctly scarce as several score of bananas were searched. At Morogoro, 
on the other hand, two males, sixteen females and sixteen young were 
taken in a small patch of bananas only two hundred yards from the 
station. As many as nine were taken beneath one leaf stalk though 
usually there were not so many. A larger series could have been 
obtained with ease. 

Megalixalus brachynemis Boulenger 

Megalixalus brachynemis Boulenger, 1896, Ann. Mag. Nat. Hist. (6), 17, p. 403, 
pi. xvii, fig. 2: Chiradzulu, Nyasaland; Loveridge, 1929, U. S. Nat. Mus. 
Bull. No. 151, p. 114: Kizerui, Usambara Mountains, Tanganyika Terri- 
tory: Ahl, 1930, Sitzber. Ges. Naturf Freunde Berlin, p. 91: Ipiana; 
Rungwe; Kilwa, etc., Tanganyika Territory. 

? Hyperolius multifasciatus Ahl, 1931, Mitt. Zool. Mus. Berlin, 17, p. 24: 
Rungwe, Tanganyika Territory. 

? Hyperolius acuticeps Ahl, 1931, loc. cit. p. 29: Ukonde-Unyika, Tanganyika 
Territory. 

Hyperolius ipianae Ahl, 1931, loc. cit. p. 43: Ipiana, Tanganyika Territory. 

Hyperolius unicolor Ahl, 1931, loc. cit. p. 122: Ipiana, Tanganyika Territoiy. 

51 (M. C. Z. 16826-36) Mwera, Zanzibar. 21. x. 29. 
314 (M. C. Z. 16837-46) Mwaya, Lake Nyasa. 1. iii. 30. 
1 (M. C. Z. 16847) Mwandemeres, Rungwe. 11. iii. 30. 
3 (M. C. Z. 16848-50) Ujiji, Lake Tanganyika. 28. v. 30. 

Affinities. As the species placed in the synonymy were based on 
single specimens, except acuticeps of which the author had two, I have 
been unable to examine the types and therefore my supposition as to 
their status is a tentative one. From the citations it will be noted that 
there are examples of brachynemis in the Berlin Museum from both 
Rungwe and Ipiana. Ipiana is ten miles from Mwaya and forty from 
Rungwe; both are in the region settled by Wakonde and Wanyika. 

Habitat. The whole series are from domestic bananas. It may be 
thought that there is some mistake in listing the whole Mwaya series 
as taken on a single day; not only is there no mistake, but all were 
brought in between the hours of 3 and 6 p.m. as described in the intro- 
duction to these reports. 

? Hyperolius sansibaricus (Pfeffer) 

Rappia sansibarica Pfeffer, 1893 (1892), Jarhb. Hamburg Wiss. Anst. 10, part 
1, p. 97, pi. ii, fig. 4: Zanzibar. 

3 (M. C. Z. 171148-50) Bagamoyo. 16. xi. 29. 



400 bulletin: museum of comparative zoology 

Affinities. Unfortunately these three frogs are damaged with rust 
from their container and are identified with grave doubts. Two of 
them have irregular tubercles on the head as described for sansiharicus 
which Boulenger once suggested was a synonym of cinctive7itris -Cope. 
In life they were unlike any other Ilyperoli with Avhich I am acquainted. 

Coloration in life. Above, bright rufous with dark brown markings. 
After chloroforming they changed to: Pale yellowish green minutely 
speckled with black, congregations of these spots forming dusky 
patches; thighs clearer, yolk-colored with narrow red (blood vessel) 
line showing through the skin, speckles just visible. Below, yellow 
tinged with pink in places, belly between fore and hind limbs satin- 
white. One of the others was drab-gray or putty-colored in life. 

Mrasurcmcnts. These three males range from 26 to 28 mm. 

Habitat. Taken on sedges in the swamp eight miles west of the town. 

Hyperolius viridiflavus (Dumeril & Bibron) 

Euchnemis viridi-flavus Dumeril & Bibron, 1841, Erpet. Gen., 8, p. 528: 
Abyssinia ( = Ethiopia). 

The five types are in excellent preservation. Each has a little pouch 
in the centre of the throat but no gular disk, if the head is strained 
backwards there is no strong fold, if slightly forwards a fold appears 
though perhaps not strong. If considered a fold it keys to viridiflavus 
if not to salinar. 

The fingers may be said to be half-webbed, actually the web extends 
to the last joint inner and outer aspects except that on the outer it 
extends as a narrow margin to the disk and on the 3rd finger it is some- 
what less on the inner side so that the 3rd finger can hardly be called 
half-Avebbed being only a third-webbed; the tibio-tarsal articulation 
of the adpressed hind limb marks the eye. The skin is smooth above 
and below but slightly granular on the flanks posteriorly. The di- 
ameter of the orbit practically equals that of the snout, it is greater 
than the distance from the anterior border of the eye to the nostril. 

Coloration in alcohol. AboAC, uniform gray except for a very few 
white specks which are practically absent from some of the series; the 
thighs are white except for a few fine specks and the trace of a narrow 
silver line. 

Hyperolius symetricus (Mocquard) 

Rappia symetrica Mocquard, 1902, Bull. Mus. Hist. Nat. Paris, 8, p. 408: 
Athi River, Kenya Colony. 



loveridge: afrkax herpktoi.ogy 401 

Hyperolius symetriciis Loveridge, 1929, U. S. Nat. Mus. Bull. 151, p. 118: 
Nairobi; Wambugu; Mt. Kenya, Kenya Colony; Loveridge, 1930, Proc. 
Zool. Soc. London, p. 28: Key to species. 

Hyperolius aspcr Ahl, 1931, Mitt. Zool. Mus. Berlin, 17, p. 49: Nairobi, Kenya 
Colony. 

In passing through Paris I availed myself of Mons. Angel's kindness 
and examined the holotype of this species which correctly responded 
to the key published in 1930 and represents the same species as the 
examples in the United States National ^Museum which were reported 
on in 1929. The following notes, confirmatory or additional to the 
original description, were made from the type. 

It is a male with gular disk and fold across the chest; the snout is 
ec^ual to the orbital diameter as stated, though on the right side the 
orbital diameter appears to only equal the distance from the nostril to 
the anterior border of the eye; the fingers are half -webbed, i.e. to the 
last articulation except the 2nd which is webbed right to the disk (or 
practically so) on the inner aspect, in this respect it differs from the 
type oi ferniquci, the outer finger is half-webbed; the webbing on the 
toes is identical with that of ferniquci , the 1st is webbed to the disk, 
the 2nd and 3rd on their outer aspect to the last joint, on their inner 
to the disk, the 4th to the last joint on the inner side, to the disk on the 
outer, the 5th to the disk. The back is smooth except for a regular par- 
allel series of dorso-lateral pimples or warts. 

Coloration in alcohol. This specimen must have been very hand- 
somely colored in life through now somewhat faded. The inter- 
orbital, sub-triangular marking is still distinct; a brown stripe from 
the nostril passes through the eye and widens into a large blotch ex- 
tending nearly to midbody ; the whole upper edge of the patch from 
the nostril is bordered by a relatively broad band of china-like white- 
ness; a similar line along the lip terminates above the fore limb, other- 
wise it gives the impression of the blotch being completely surrounded 
by white; another blotch on the flank is three-quarters surrounded 
by a white edge; there are similar white-edged blotches on the thigh, 
tibia and above the anus which continue and complete the pattern 
when the frog is at rest and thus assist in breaking up its outline. 
Below, white, the throat finely speckled with brown, a few brown 
speckles on the abdomen and lower sides. 

Hyperolius striolatus Peters 

Hyperolius striolatus Peters, 1882, Sitzber. Ges. Naturf. Freunde Berlin, p. 9: 
Taita, Kenya Colony. 



402 bulletin: museum of comparative zoology 

Rappia ferniquei Mocquard, 1902, Bull. Mus. Hist. Nat. Paris, 8, p. 407: Athi 

River, Kenya Colony. (Atchi error e). 
Rappia marmorata Procter {nee. Rapp), 1920, Proc. Zool. Soc. London, p. 417. 
Hyperolius marmoratus Loveridge, (part), 1929, U. S. Nat. Mus. BuU. No. 151, 

pp. 116-7: various localities in Kenya. 
Hyperolius coeruleopunctatus AM, 1931, Mitt. Zool. Mus. Berlin, p. 76: Kibwezi 

and Nairobi, Kenya Colon}'. 
Hyperolius udjidjiensis Ahl (part), 1931, loc. cit. p. 97: Kibwezi, Kenya Colony. 

(The Ujiji specimen may be distinct). 

The Museum of Comparative Zoology has examples of this frog from 
Bissel, XairoVji and Xaivasha in Kenya Colony. Whether any constant 
morphological characters will be found to distinguish the somewhat 
similar forms from the Central Lake Region (Ujiji, Lake Tanganyika; 
Kissenje, Lake Kivu; Bukoba, Lake Victoria, etc.) remains to be seen. 

The Kenya frogs are certainly identical with striolatns as figured by 
Ahl (1931, Das Tierrich, 55, pp. 312-3, fig. 187) and when the time 
comes for a revision of the genus, striolatns will certainly have to be 
recognized as distinct from marmoratus though possibl}^ only as a 
color form. In 1920, the late Miss Procter, following Boulenger who 
considered striolatns a synonjTn of marmoratus, referred my Nairobi 
frogs to the latter species ; tentatively and with reservations I followed 
this procedure in 1929 {loc. cit.) when dealing with the Smithsonian 
Expedition material: the color varieties numbered 2, 3 and 4 in that 
paper should be referred to striolatus. 

The holotype oi ferniquei was another of the frogs which I was able 
to examine when in Paris with interesting results. Most important of 
these was the discovery of my own error (1930, Proc. Zool. Soc. 
London, p. 25) in stating that the tympanum of this species was dis- 
tinct, whereas it is hidden as in all other East African members of the 
genus. The error arose at the time I was stationed in Nairobi when a 
good friend sent me translations of certain descriptions which were not 
procurable in East Africa. In translating the description of ferniquei 
he wrote " distinct" instead of " indistinct," and I never had occasion 
to check this translation until recently. 

The holotype of ferniquei is a male with a gular disk and a slight fold 
behind it. The diameter of the orbit is equal to the distance from the 
anterior border of the eve to the nostril, not to the end of the snout ; 
the fingers are half-webbed, i.e. to the first joint, the outermost finger 
is half-webbed but not more; the toes are beautifully preserved and 
when seen from below the 1st is webbed to the disk (alternatively it 
might be said to be webbed halfway between the terminal joint and the 



loveridge: African herpetology 403 

disk and continued to the latter as a narrow margin), the 2nd and 3rd 
to the disk on the inner side, to the last joint on the outer, the 4th to 
the last joint on the inner and to the disk on the outer, the 5th to the 
disk. The back is smooth though by careful search a slight granulation 
is apparently to be observed. ^Yhen tested by my 1930 key to the 
species of Ilyperolius, the type oi ferniquei agrees with m armor aivs if 
one considers that it has a strong fold across the chest, if no strong fold, 
then it runs down to salinae. 

Its present color in alcohol is a slightlj^ pinkish buff vermiculated 
with brown all over the upper surface, the vermiculations being formed 
of many juxtaposed fine dots; on the flanks these dots are separated 
and blacker. Below, white, the throat finely dotted all over, the dots 
coalescing to form two dark patches situated posterio-laterally. It is 
undoubtedly much faded. 

Hyperolius callichromus Ahl 

Hyperolius giiUulatus Barbour & Loveridge {nee. Giinther), 1930, in Strong, 

African Republic of Liberia, 2, p. 794: Uvira, Lake Tanganyika, Belgian 

Congo. 
HyperoU-us callichromus Ahl, 1931, Mitt. Zool. Mus. Berlin, p. 99: Western 

bank of Rusizi River and Northwest shore of Lake Tanganyika, Belgian 

Congo. 

21 (M. C. Z. 17151-60) Nyamkolo, Lake Tanganyika. 9. v. 30. 
1 (M. C. Z. 17161) Ujiji, Lake Tanganyika. 28. v. 30. 

Affinities. The coastal frogs (Dar es Salaam, Bagamoyo, etc.) 
referred to callichromus by Ahl are more probably referable to punc- 
ticulatus Pfeffer which occasionally produces mutants similar in dorsal 
pattern to callichromus; as far as my experience goes, however, such 
coastal frogs may be distinguished from callichromus of the Lake Re- 
gion by their retention of a streak connecting the nostril and eye. 

The frog from Uvira which is at the extreme north end of Lake 
Tanganyika just west of the mouth of the Rusizi River is undoubtedly 
conspecific with callichromus being an example in which the vertebral 
streak has broken up into a series of spots, similar individuals occur in 
the Nyamkolo series with which it also agrees in having large blotches 
or streaks on the tibia. On the other hand two cotypes of callichrovms 
(M. C. Z. 17630-1) differ from the Uvira and Nyamkolo frogs in having 
the tibia finely speckled in place of one or more large blotches. The 
Ujiji frog combines both types for it possesses both large blotches and 
fine speckling. 



404 bulletin: museum of comparative zoology 

Coloration in life. The whole series is irregularly spotted and 
streaked. The actual coloring has several times been figured under 
the name of marmoraius but the Central African frog appears to be of 
larger size than the Natal species. 

Measurements. The largest frog, a female, measures 36 mm. 

Breeding. Females from both Nyamkolo and Ujiji hold well-devel- 
oped ova. 

Diet. Beetles and a hemipteron. 

Habitat. Found squatting on sedges growing from deep water. They 
are very conspicuous but also exhibit unusual activity for frogs of this 
genus. When approached they show a tendency to leap away before 
one is at all close; they then dive and either remain below the surface 
for some time, or swim away to a distance before coming to the surface. 

Hyperolius rhodoscelis (Boulenger) 

Rappia rhodoscelis Boulenger, 1901, Ann. Mus. Congo, (1) II, fasc. 1, p. 3, pi. 
ii, fig. 1 : Pweto, Lake Mweru, Belgian Congo. 

41 (M. C. Z. 17236-50) Nyamkolo, Lake Tanganyika. 9. v. 30. 

Distribution. Formerly knoAvn only from the type locality which is 
thirty miles due west of Nyamkolo. 

Affinities. Boulenger has pointed out that this species is nearly 
related to //. marmoraius and //. argus; it seems to be even nearer to 
picturatus, from which it differs in its blunter snout and larger size. 

Variation. The interorbital space, said to be " pen plus etroitc'' than 
the upper eyelid is often as much as twice as broad. The throat is 
granular in males but not in the females. 

Coloration in life. Adult. Above, rich green (or chrome yellow), a 
purplish brown line, composed of closely juxtaposed dots, commences 
at the nostril, passes through the eye, broadens just behind the eye and 
then diminishes along the flank to the thigh where it merges into an 
area of widely separated dots on the thigh ; a narrow vitta of green 
(or yellow) on the upper surface of the thigh, anterior surface of the 
thigh vermilion (not blood red) as is also the posterior in the immediate 
vicinity of the knee, the vermilion coloring spreading on to the tibia; 
upper surface of the foot yellow (or orange) splashed with vermilion. 
Below, white on throat; cream on breast and belly; colorless on limbs; 
orange on soles of hands and feet. 

Young. Above, a dusky, satiny brown composed of a multitude of 
fine specks; a light, dark -edged vertebral stripe commencing between 
the eyes, extends to the anus; it is flanked on either side by similar but 



loveridge: African herpetology 405 

slightly broader, light lines commencing at the nostrils, passing 
through the eyes and along the flanks to terminate at the groin; a less 
well-defined line still lower on the flanks is apt to merge into the white 
of the belly; a splash of vermilion on the anterior surface of the thigh 
which is uniformly speckled with greyish brown above, on the tibia 
these dots tend to form longitudinal stripes; hands and feet yellowish. 

During development the white lines of the young become yellowish, 
then yellow, and spread till they merge to form the uniform yellow 
(eventually green) back; the dark edging on the lower side of the upper- 
most lateral band broadens and deepens in color to form the side 
streak of the adult frog. The livery of the young appeared to me to be 
indistinguishable from that of young pundiculatus from the Uluguru 
Mountains. That they are the young of rhodoscelis, however, seems 
obvious despite a trifling difference in the amount of webbing. Both 
young and adult were taken in the same patch of sedges. 

Measurements. Eight males range from 23 to 31 mm., average 27 
mm.; twenty females range from 23 to 34 mm., average 29 mm. due 
to so many immature specimens; thirteen young range from 18 to 22 
mm., average 22 mm. The gular disk of the male is apparent at 23 mm. 

Breeding. While some females were bloated with eggs, others had 
evidently laid; some of the numerous young still showed tails when 
measuring 18 and 19 mm. from snout to anus, though others of the 
same dimensions had lost their tails. 

Diet. Most stomach contents were too finely masticated for identi- 
fication, the following, however, were recognisable: (1) large orthop- 
teran, (2) caterpillar, (3) bug, weevil and skipjack beetle, (4) minute 
neuroptera, spiders. 

Hyperolius picturatus Peters 
Hyperolius picturatus Peters, 1875, Monatsb. Akad. Wiss. Berlin, p. 206, pi. ii, 
fig. 2: Boutry, Ashanti, Gold Coast. 

3 (M. C. Z. 17251-3) Kampala, Uganda, vi. 30. 

Distribution. This species has often been recorded from Uganda 
and is reported from Kenya Colony and Pemba Island. This last 
record of Boettger's to judge by his remarks and comparison with 
Tornier's figure 108 of plate ii, makes it practically certain that he had 
a specimen of H. puncticulatus (Pfeffer). 

Variation. These frogs have been compared with specimens from 
the Gaboon (det. Boulenger) and Cameroon and appear to be speci- 
fically identical. 

Measurements. The males measure 24 and 27 mm., the female 28 mm. 



406 bulletin: museum of comparative zoology 



Hyperolius puncticulatus (Pfeffer) 

Rappia puncticulata Pfeffer, 1893, Jahrb. Hamburg Wiss. Anst. 10, p. 99, pi. ii, 
fig. 2: Zanzibar. 

Rappia argus Procter (nee. Peters), 1920, Proc. Zool. Soc. London, p. 417: 
Morogoro and Dar as Salaam, Tanganyika Territory. 

Hyperolius argus Barbour & Loveridge (nee. Peters), 1928, Mem. Mus. Comp. 
Zool., 50, p. 222: Nyingwa and Vituri, Uluguru Mtns., Tanganyika Terri- 
tory. 

Hyperolius subsiriaius Ahl, 1931, Mitt. Zool. Mus. Berlin, 17, p. 84: Amani, 
Dar es Salaam, Uhehe, Ukonde-Unyika, Ujiji, etc., Tanganyika Territory. 

Hyperolius calliehromus Ahl (part), 1931, loc. cit. p. 99: Bagamoyo, Dar es 
Salaam, etc., Tanganyika Territory. 

1 (M. C. Z. 17162) Mwera, Zanzibar, 21. x. 29. 
1 (M. C. Z. 17163) Mwaya, Lake Nyasa. 1. iii. 30. 
3 (M. C. Z. 17164-6) Ilolo, Rungwe. iii. 30. 

Distribution. The Zanzibar frog is a topotype. Having collected 
large series of this frog at Amani, Dar es Salaam, etc. on previous 
expeditions no particular attempt was made to get them on the 
present one. Ilolo is almost on the Unyika Plateau from which the 
British Museum has a series of these frogs. The distribution of this 
species appears to be dependent upon banana plantations and is little 
affected by altitude for they occur from the coast up to 7,500 feet. 

Affinities. Parker (1931, Proc. Zool. Soc. London, p. 902) has re- 
cently shown that the specimens from the Uluguru Mountains referred 
by Barbour and myself to argus, are really variants of pu7icticulatus. 
The misidentification was due to the fact that our only comparative 
material (Xyika Plateau, det. Boulenger, and Morogoro, det. Procter) 
was also pundiculatus misidentified as argus. Whether the Ujiji 
cotype of suhstriatus Ahl is conspecific is questionable. 

Coloration in life. In the field I noted that the Ilolo frogs were 
brought in with a series of marginatus Peters from which they were 
morphologically indistinguishable though undoubtedly not specifically 
identical; they were separated on the basis of the cantho-lateral band 
which is typical of puncticulatus. 

Measurements. The largest frog, a female, measures 34 mm. 

Hyperolius marginatus Peters 

Hyperolius marginatus Peters, 1854, Sitzber. Akad. Wiss. Berlin, p. 627: 
Maganga, Mozambique; 1882, Reise nach Mossamb., 3, p. 165, pi. xxii, 
fig. 8. 



loveridge: African herpetology 407 

Hyperolius pictus Ahl (part), 1931, Mitt. Zool.Mus. Berlin, 17, p. 44: Ngosi 
Volcano Crater Lake; Uhehe; Iringa; Rungwe; Ukinga Mountains, 
Tanganyika TerritorJ^ 

Hyperolius ngoriensis Alil, 1931, loc. cit. p. 60: Ngosi Volcano Crater Lake, 
Tanganyika Territory. 

Spawn and 38 (M. C. Z. 17174-84) Dabaga, Uzungwe Mtns. 1. i. 30. 

1 (M. C. Z. 17276) Boma Ngombe, Uzungwe Mtns. 4. i. 30. 
14 (M. C. Z. 17185-93) Kigogo, Uzungwe Mtns. 13-30. i. 30. 

1 (M. C. Z. 17194) Lukungu, Uzungwe Mtns. 8. ii. 30. 

2 (M. C. Z. 17195-6) Ihenye, Uzungwe Mtns. 8. ii. 30. 
1 (M. C. Z. 17197) Mangoto, Ukinga Mtns. 10. ii. 30. 

3 (M. C. Z. 17198-200) Madehani, Ukinga Mtns. 19-21. ii. 30. 
122 (M. C. Z. 17201-225) Nyamwanga, Poroto Mtns. 17. iii. 30. 

25 (M. C. Z. 17226-35) Ngosi Crater, Poroto Mtns. 18. iii. 30. 
7 (M. C. Z. 17167-74) Ilolo, Rungwe District. 19-30. iii. 30. 

Distribution: I have taken the liberty of correcting the misspelhngs 
of the type localities as given by Ahl, viz. Ngosi for " Ngori," Rungwe 
for "Rugwe," Unyika for "Nika." 

I very much doubt if the paratypes of pictus from Nairobi, Bukoba 
and perhaps some of the other localities are specifically identical with 
those from the Ngosi Crater. 

Native names. Kolamivihve (Kihehe); tufi (Kikinga, probably not 
specific or generic). 

Affinities. This species appears to be related to concolor Hallowell 
and picturatus Peters of West Africa; several of Bocage's Angolan 
species may be synonymous. 

That the 227 frogs listed above are all one species there is not the 
slightest doubt; nor is there the least question of their specific identity 
with Ahl's pictus and ngoriensis which are figured on pages 302 and 324 
of Das Tierrich, 1931, 55, Amphibia Anura iii. Five of the variations 
are very well portrayed in fig. 176 (page 302) and were observed in the 
field where all the above material was provisionally listed and identified 
as one species. 

It is possible, though improbable, that marginatus Peters is distinct 
for I have no topotypical material for comparison. The coloring of 
Peter's figure in Reise nach Mossambique is presumably hypothetical 
and incorrect though the pattern is accurately delineated. Moreover 
Ahl records the distribution of marginatus as "Mozambique bis 
Deutsch-Ostafrika," obviously by the latter he intended Tanganyika 
Territory. 

It will be seen that the above series contains topotypes of all three of 
Ahl's " species " for 7igoriensis is only a young pictus and the figure is in 



408 bulletin: museum of comparative zoology 

complete agreement with juveniles from Ngosi Crater in the present 
series. 

Coloration in life. The following notes were made on catching the 
first examples of this frog as it was recognised as a species never before 
encountered by me in Kenya Colony or Tanganyika Territory: 

Dabaga. Color very variable, hardly two alike but in all green pre- 
dominates above, yellow below and there is blood red on the hinder side 
of thighs and tibia. Above, olive, a yellow green streak from nostril 
over the eye nearly to the groin, two straight, greenish, dorso-lateral 
streaks within these lines which break up into yellow, black-edged, 
spots and streaks in the vicinity of the anus; fore and hind limbs olive, 
spotted with yellow green, blood red on the hinder sides of thigh and 
tibia. Below, bright chrome yellow. 

Lukungu. A female whose back was dark sap-green. Below yellow- 
ish white. 

Ihenye. Above uniform gamboge, a very irregular dark line as of 
diamond-shaped areas united by parallel lines from the nostril through 
the eye to the flank where it breaks up and disappears about midbody. 
\Yhen the hind limbs are in a position of rest it will be seen that the 
line is continued upon them (very broadly on the thigh which is hidden 
at rest) narrowly on the tibia and foot. The inside angles between the 
thigh and tibia, tibia and tarsus, and the top of the foot are blood red. 

Mangoto. A female with the same coloring as the last but with 
more brown mottling and marbling on the hind limbs. 

Kigogo. A frog with green stripes was placed in a vivarium with 
dead, brown grass, whereupon it turned uniformly olivaceous brown. 

Nyamwanga. It was noted that the color pattern of the young 
shows great variability. 

Measurements. A large 9 (M. C. Z. 17189) measures 34 mm., but 
this is unusual from 26 to 29 mm. being a more usual size ; several cf cf 
are 29 mm. 

Breeding. AtDabagaspawn,ofwhatwas almost certainly this species, 
was collected on January 1st; the only other frogs found in the vicinity 
were males of Rana ansorgii and both sexes of Arthroleptis minutus. 

At Kigogo, on January 30th, four frogs were found in, or on the edge 
of, two large puddles on the path. A very small male was embracing 
a big female, the latter deposited black and white eggs (each more or 
less enveloped in an independent gelatinous sphere, though the spheres 
were not detached), in the water dish of the vivarium on February 1st. 

Diet. Hemipteron, beetles, ants, a wasp and a spider were found in 
stomachs of this species. 



loveridge: African herpetology 409 

Eneviies. A Nyamwanga frog had lost a hind leg just below the 
knee-joint which had healed up. 

Habitat. The majority were taken in heavily grass-grown, swampy 
ground in \alleys of the mountains; at Kigogo one was found squatting 
in a shrub; the Madehani series were taken beside a stream in an open 
valley ; some of the Ngosi Crater series in wild bananas. 

Hyperolius mariae Barbour & Loveridge 

HyperoUus mariae Barbour & Loveridge, 1928, Mem. Mus. Comp. Zool., 50, p. 

217, pi. iii, fig. 1: Derema, Usambara Mtns., Tanganyika Territory. 
Hyperolius fuelleborni Ahl, 1931, Mitt. Zool. Mus. Berlin, 17, p. 75: Langen- 

burg (i.e. Manda) on Lake Nyasa; Rungwe etc., Tanganyika Territory. 

7 (M. C. Z. 17254-60) Mwaya, Lake Nyasa. 1-8. iii. 30. 

Native name. Korfe (Kinyakusa). 

Affinities. It was no small surprise to find these frogs at Mwaya 
and it appears probable that with such a wide distribution the species 
will eventually be synonymised with some earlier form. There is no 
doubt in my mind as to the correct identification, they differ from 
concolor in the absence of pigmentation on the thighs. Males have a 
gular disk and females a strong gular fold and pouch in the middle of 
the fold. A cotype of fuelleborni has been compared with them and 
with the type of mariae. 

Coloration in life. In the field two of these frogs were compared with 
the colored plate of mariae with which they closely agreed. There is, 
however, wide variation in the dorsal markings of the other five which 
range from minute stippling with black specks, through small black 
spots to one which possesses large blotches on the back. 

Measurements. The larger cf measures 29 mm., the largest of five 
9 9 30 mm. 

Breeding. The largest females hold well-developed ova. 

Habitat. Taken on sedges in a swamp as at Derema; the climate at 
Mwaya is much hotter, however. 

? Hyperolius platyrhinus (Procter) 

Rappia platyrhinus Procter, 1920, Proc. Zool. Soc. London, p. 416, text-fig. 3: 
Nairobi, Kenya Colony. 

1 (M. C. Z. 17261) Shinyanga, Usukuma. 3. vi. 30. 
1 (M. C. Z. 17262) Jinja, Uganda. 3. vii. 30. 

Affinities. I am far from satisfied that these two frogs are specifically 



■ilO bulletin: museum of comparative zoology 

identical with platyrhinus but the Shinyanga frog is certainly the same 
as a series of twenty-five frogs from Xyambita, Usukuma, due north of 
Shinyanga, which I referred to platyrhinus in 1925 (Proc. Zool. Soc. 
London, p. 785). 

Measuremenis. These males measure 25 and 22 mm. respectively. 

Coloration. The thighs of both are minutely speckled with black as 
are the backs. 

Aestivating. The Shinyanga frog has extensive deposits of fat. It 
was found on the door of a safe standing on a verandahof ahouseouton 
a plain with no trees or shade within several hundred yards. The owner 
of the house said that it was on the safe when he left home a month 
before and was still there on his return on June 3rd at which time the 
weather was very hot. 

Hyperolius granulatus (Boulenger) 

Rappia granulala Boulenger, 1901, Ann. Mus. Congo, (1) ii, fasc. 1, p. 4, pi. ii, 
fig. 3: Pweto, Lake Mweru, Belgian Congo. 

2 (M. C. Z. 17274-5) Nyamkolo, Lake Tanganyika. 9. v. 30. 

Distribution. The type locality is thirty miles due west of Nyamkolo. 
There is an example of this frog from "Kinyamkolo" in the British 
Museum labelled H. nasida. I have compared my specimens with a 
true nasuta from Ngola, Angola but that species has a much sharper 
snout. 

Variation. Except with a lens, the granular nature of the skin is not 
noticeable in these formalin-preserved specimens. 

Coloration in life. Rich green; unfortunately no notes were taken 
in the field as it was mistaken for a form of microps. 

Measurements. Both are females and measure 19 and 21 mm. 

Breeding. Both distended with ova. 

Habitat. Taken among sedges. 

Hyperolius parkeri sp. nov. 

Hyperolius microps (part) Barbour & Loveridge (not of Giinther), 1928, Mem. 
Mus. Comp. Zool., 50, p. 225, and Loveridge, 1932, Proc. Biol. Soc. 
Washington, 45, p. 63. 

19 & eggs (M. C. Z. 17263-73) Bagamoyo. 16. xi. 29. 

Type. Museum of Comparative Zoology, No. 13365. An adult 9 
from Mogogoni swamp, south of Dar es Salaam, Usaramo, Tanganyika 
Territory, collected by Arthur Loveridge, November 10, 1926. 



loveridge: African herpetology 411 

Paratypes. Museum of Comparative Zoology, Xos. 1336G-7 from 
Derema, Usambara Mountains, Tanganyika Territory, collected 
30. xi. 26, and nineteen others from Bagamoyo as listed above. 

Correction. The five frogs referred to microps Giinther by Barbour 
and Loveridge in 1928 have subsequently proved (as a result of the 
capture of the Bagamoyo series comprised of both sexes) to represent 
two species, both of which occur at Dar es Salaam. A pair of these have 
become the types of H. usaramoae Loveridge (1932, Proc. Biol. Soc. 
Washington, 45, p. 63) but the others, which at that time I still thought 
must represent microps, are not that species according to Parker. 
The frogs referred to as "microps" in that paper are, therefore, the 
types of parkeri, so named in appreciation of Mr. Parker's kindness in 
comparing the type (M. C. Z. 13365) with the type of inicrops. 

Diagnosis. The type, having been compared wdth the type of 
microps Mr. Parker writes: "Your specimen 13,365 is twice as big as 
the type of microps with a longer, flatter snout, appreciably more web 
between the toes, and a different color pattern." Parker's statement 
that the Dar es Salaam frog is twice the size of microps reveals the fact 
that when Giinther stated the length of microps to be ten lines, he must 
have used line in the metric sense and not in the more usual English 
definition as a twelfth of an inch. In 1928 I assumed the latter inter- 
pretation when stating that the type of microps was 21 mm. in length, 
apparenth' it is 10 mm. 

H. parkeri differs from usaramoae precisely in the way that the 
latter is stated to differ from "microps'' in the 1932 paper where one 
should substitute "parkeri" for "microps." 

H. parkeri differs from peter si Ahl (only known to me from the 
description and figure) in the head being markedly longer than broad, 
the more acuminate snout, the upper jaw projecting beyond the lower 
and the more extensive webbing of the toes. 

Males in the Bagamoyo series are distinguished from the females, 
not only by then- granular gular disk but by the spines (probably only 
a breeding season development) on the soles of their feet. Such spines 
being absent in the females. 

Coloration in life. Type 9 , gravid, numerous ova observable 
through the semi-transparent abdominal skin. Above, rich green, 
a silvery lateral line (disappears on preservation) bordered above and 
below by a series of deep black spots; the whole upper surface is 
speckled with minute black specks; limbs greenish yellow. Below, 
a faintly greenish shade tinged with blue on the throat. 

In the Bagamoyo series the sexes exhibit a marked color difference. 



412 bulletin: museum of comparative zoology 

The males are brown, or olive, exactly like the tints of the dead and 
dying sedges while the females were of a vividly fresh green hue like 
the living sedges. This green pigmentation is partly soluble, for the 
water in which dead parkeri are soaked for a few hours, takes on a 
green tinge. The throats of the frogs of either sex were scarcely blue, 
certainly not so pronouncedly so as in the type, but as the Bagamoyo 
series were collected six days later in the month and had already de- 
posited ova, it may be that the blue disappears after oviposition. The 
light lateral line, marked above and below by parallel series of black 
dots, is present in both sexes. 

The thighs of the males show some pigmentation and are not 
colorless like those of usaramoar males. 

Measurements. Type 9 . Head and body to anus 23 mm., breadth 
of head 7.5 mm., length of head to angle of jaw 8 mm., length of snout 
from nostril 1.5 mm., length of snout from anterior border of orbit 
4 mm., length of hind limb from anus 40 mm., length of fourth toe 
5.5 mm. 

The following data is based on the Bagamoyo series which, unfor- 
tunately, are stained by rust resulting from the corrosion of their con- 
tainer penetrating the wrappings during many months at the coast 
awaiting shipment. Five males range from 20 to 22 mm.; eleven 
females from 19 to 22 mm., average for both sexes being 20 mm.; 
three young measure 11, 12, and 19 mm. respectively. 

Breeding. These frogs were calling from sedges in a swamp six miles 
out of town on the Ngerengere Road. The call was ringing and clear 
like a "pop-pop." The sedges were so sharp that it was almost im- 
possible for my bare-legged assistants to get through them. Fortu- 
nately I was wearing rubber hip-boots and by advancing towards the 
boys with the broad length of my foot forward so as to trample down 
the sedges, I was able to drive the frogs, which were squatting on the 
sedges just a few inches above the water level, before me until they 
were caught by the boys The water was just a foot deep. 

The white eggs are laid on the sedges just above the water level; 
if the rains continue as they should do, then the eggs would be sub- 
merged in the course of a few days. The eggs, embedded in an oval 
patch of colorless jelly, numbered about sixty-nine and a hundred 
and ten respectively, these hatches representing the laying of two 
frogs. 

Kassina senegalensis (Dumeril & Bibron) 

Cystignalhus senegalensis Dumeril & Bibron, 1841, Erpet. Gen., 8, p. 418: Lakes 
in the vicinity of Galam, Senegal. 



loveridge: African herpetology 413 

1 (M. C. Z. 16775) Mainland opposite Kilindini. 29. x. 29. 
1 (M. C. Z. 16776) Mwaya, Lake Nyasa. 1-8. iii. 30. 
1 (M. C. Z. 16777) Mwandemeres, Rungwe. 11. iii. 30. 
26 (M. C. Z. 16778-99) Ilolo, Rungwe. 15. iii. 30. 
1 (M. C. Z. 16800) Ukerewe Id., Lake Victoria. 10. vi. 30. 

Native name. Dori/a (Kinyakusa). 

Affinities. Hewitt (192(j, Ann. S. African Mus., 20, p. 488) has 
reinstated K. wealii Boulenger as a S. African species. The material 
listed above consists of twenty-two males and eight females and 
upholds his definition of senegalensis. 

Measurements. The males range from 33 to 42 mm., average 38 mm. ; 
the females range from 29 to 42 mm., with an average of 37 mm, 
though the breeding females (i.e. Ilolo series) average 40 mm. 

Breeding. The Ilolo series were brought into camp by two small 
boys who had found them in a pool; evidently the males assemble first 
as they outnumbered the females by three to one. These females, as 
also the one from Mwaya, were distended with ova. 

Habitat. Owing to its subterranean habits this frog is rarely en- 
countered except when the rains break and they assemble to breed. 
The immature female from Kilindini was dug out of a male king- 
fisher's resting burrow in a sandy bank. The Ukerewe frog is also a 
young female and was found by Salimu beneath a log at the edge of a 
patch of dry bush. 

BREVICIPITIDAE 

Breviceps mossambicus Peters 

Breviceps mossambicus Peters, 1855, Arch. Naturg., 21, part 1, p. 58: Island of 
Mozambique and Sena. 

8 (M. C. Z. 16430-7) Masiliwa, Turu. 10. xii. 29. 
2 (M. C. Z. 16438-9) Mangasini, L^sandawi. 14. xii. 29. 
24 (M. C. Z. 16440-50) Ilolo, Rungwe. 15-30. iii. 30. 

Distrihntion. Recorded from Iringa and Ipiana by Nieden. 

Native name. Tuvye (Kinyakusa, but not specific). 

Measurements. The largest, a Masiliwa frog, measures 52 mm., three 
of the Ilolo series are very small being 19, 22 and 26 mm. respectively. 

Diet. Termites, both at Masiliwa and Ilolo. 

Enemies. One was recovered from the stomach of a Boomslang 
(Dispholidus typus) at Masiliwa. 

Habitat. The eight Masiliwa frogs were taken as they hopped away 



414 bulletin: museum of comparative zoology 

from the woodland path along Avhich I was cycling in the early morning 
after a night of heavy rain which ushered in the rainy season in this 
district. Most of the frogs were on patches of sodden leaves and none 
were seen after 9 a.m. 

Probreviceps macrodactylus rungwensis Loveridge 

Probreviceps macrodactylus rungwensis Loveridge, 1932, Bull. Mus. Comp. 
Zool., 72, p. 387: Nkuka Forest, Rungwe Mtn., southwestern Tanganyika 
Territory. 

4 (M. C. Z. 16451-4) Nkuka Forest, Rungwe Mtn. 25-30. iii. 30. 

Native name. Tuvye (Kinyakusa but also applied to Kassina, 
Phrynomerus and Breviccps). 

Breeding. Ova are moderately developed in the females. 

HoPLOPHRYNE ULUGURUENSis Barbour & Loveridge 

Hoplophryne uluguruensis Barbour & Loveridge, 1928, Mem. Mus. Comp. 
Zool., 50, p. 254, pi. ii, figs. 3 and 4: Nyange, Uluguru Mtns., Tanganyika 
Territory; Noble, 1929, Bull. Am. Mus. Nat. Hist., 58, pp. 291-313. 

In his paper on " The Adaptive Modifications of the Arboreal Tad- 
poles of the genus Hoplophryne," Noble makes several statements 
which are hardly in conformity with my published field notes which 
are the only source of information on the habits of this frog. 

On page 292, the statement that H. uluguruensis " lays its eggs be- 
tween the leaves of the wild banana" is a misconception which appears 
in one form or another at many places though rightly stated under 
"Conclusions" on page 330 as "between leaf and stalk of banana 
plants." 

The rainwater is retained between stalk and leaf stalk, not " between 
the leaves of the wuld bananas." 

I cannot concur with the statement that " It is also possible that 
some of the tadpoles in the banana leaves may not reach the pockets 
of water but live exposed to air as do the tadpoles in the bamboos." I am 
not aware that anyone has yet found the tadpoles alive in bamboos, 
it is certainly highly improbable that they are not submerged in the 
water retained by the internode. Though no mention was made of 
water being held by these internodes under the heading of //. ulugu- 
ruensis in the 1928 paper, remarks had been made earlier on in this 
paper under Nectophrynoidcs spy. which were taken at the same spot. 
If one could visualize the heavy, driving rain storms of daily occur- 



loveridge: African herpetology 415 

rence in October, and the continual mist -drenched condition of the rain 
forest at 7,500 feet where the bamboos were growing, it is difficult to 
conceive how any receptacle capable of retaining water was not so 
doing. The suggestion, or theorj', that these tadpoles do not live in 
water is apparently based on an interpretation of their respiratory 
adaptations and pulmonary structure for on page 300, Noble suggests 
" that the larvae do not swim in the water caught between the banana 
leaves but remain on the edges of it or merely wriggle in damp crannies 
between the leaves. This conclusion is supported by other evidence 
to be discussed below." Apparently my statement that all these tad- 
poles were found swimming in the water is of no consequence and no 
reference is made to it. Even if it does nullify much attractive theoriz- 
ing it seems advisable to repeat that, with the exception of tadpoles 
observed hatching and wriggling down the moist surface of the stem 
into the water, in no single instance ivas a live tadpole found any^vhere 
hut in the ivafer retained betiveen stalk and leaf stalk and tadpoles were 
found in about a score of the banana plants examined. It is difficult 
to imagine one of these tadpoles ever wriggling its way up the wet and 
slippery vertical stem of a banana plant within the outer leaves though 
one must admit that the adult frogs accomplish the feat. It is still less 
possible for the tadpoles to make their way from one pocket of water 
to another so that such statements as that on page 306 to the effect 
"that the larvae wriggle about in the crevices at the base of banana 
leaves and pick up with their toothless but powerful jaAvs what bits of 
animal or plant debris may occur there," appears to be somewhat idle 
speculation if intended to convey the impression, as is apparently the 
case, that these tadpoles are not wholly aquatic. 

Phrynomerus bifasciatus (Smith) 

Brachymerus bifasciatus A. Smith, 1849, lUus. Zool. S. Africa, 3, pi. Ixiii: 
Country to the east and northeast of Cape Colony. 

24 (M. C. Z. 16410-25) Mwaya, Lake Nyasa. 1. iii. 30. 
4 (M. C. Z. 16426-9) Mwandameres, Rungwe. 11. iii. 30. 

Distribution. Xieden has recorded this species from Tanga, Baga- 
moyo, Dar es Salaam, Tukuyu and Ipiana near Mwaya. 

Native name. Tuvye (Kinyakusa, but not specific). 

Variation. There is no question that these frogs are typical bifas- 
ciatus and not P. affinis Boulenger from Lake Mweru. The digits and 
toes are long and slender and the terminal expansions larger than in 



416 bulletin: museum of comparative zoology 

Kilosa specimens, much depends apparently on methods of preserva- 
tion as to the degree of expansion. 

Measurements. The largest, a Mwandemeres frog, measures 58 mm. 

Diet. Ants. 

Defence. I have previously drawn attention to the poisonous nature 
of the secretions of this frog, a further example came to my notice at 
Mwaya. One of my boys brought me a bag containing a mixed catch 
of frogs from bananas — Hyperolius, Megali.valus, LeptopcHs and half- 
a-dozen Phrynomerus. I chloroformed the whole lot in the bag. An 
hour later I tipped the catch out on to a table and began picking out 
the various species. The Phrynomerus had exuded a considerable 
amount of intensely sticky dermal secretion which had gummed the 
smaller Megalixalus together. After separating these and dropping 
them into water I could not get the gummy mucous off my fingers by 
washing and so rubbed them in the dust — as a monkey would do — 
then by rubbing them together shed the mucous like so much gutta- 
percha. Shortly afterwards irritation set in on my finger tips entirely 
comparable to the irritation produced by stinging nettles and it 
actually appeared to spread within my arm up to the elbow of the right 
arm; perhaps I should add that the weather was very hot and my pores 
probably wide open. 

Habitat. All were taken in domesticated banana plants. 



EXPLANATION OF PLATES 



PLATE 1 



LovERiDGE. — African Herpetology 



PLATE 1 

Fig. 1. Lygodadylus picturatus piduratus (Peters), cf (M. C. Z. 30511) Dar es 
Salaam, Tanganyika Territory. 

Fig. 2. Lygodadylus piduratus var., cT (M. C. Z. 30590) Kilindini, Mombasa 
Island, Kenya Colony. 

Illustrating the two color phases occurring on the East African coast 
north of Tanga. Both are found on Mombasa Island though not together. 
Both enlarged about Ij^ natural size. 

Total length of each gecko 81 mm. 



BULL. MUS. COMP. ZOOL. 



LovERiDGE. African Herpetology. Plate 1 




^ 



THE MERIDEN GRAVURE CO. 



PLATE 2 



LovERiDGE.— African Herpetology 



PLATE 2 

Fig. 1. Agama agama turuensis Loveridge, type cf (M. C. Z. 30686) Turu, 
Tanganyika Territory. 

Fig. 2. Agama agama dodomae Loveridge, topotype cf (M. C. Z. 30739) 
Dodoma, Tanganyika Territory. 

Fig. 3. Agama agayna ufipae Loveridge, type d"- (M. C. Z. 30741) Kipili, 
Ufipa, Tanganyika Territory. 

Fig. 4. Agama agama mwanzae Loveridge, cf (M. C. Z. 30648) Mwanza, 
Tanganyika Territory. 

All enlarged about Ij^ natural size. 

Showing the gular coloring characteristic of four races of rock-dwelling 
agamas. The throat is displayed by the courting male in an upward jerk. 
Females lack the gorgeous coloring of the males. 



BULL. MUS. COMP. ZOOL. 



LovERiDGE. African Herpetology. Plate 2 




sa-i. 





THE MERIDEN GRAVURE CO. 



PLATE 3 



LovERiDGE. — African Herpetology 



PLATE 3 

Fig. 1. Amphisbaena mpwapwaensis Loveridge, type cf (M. C. Z. 30767) 
Mpwapwa, Ugogo. 

Fig. 2. Zonurus ukingensis Loveridge, type cf (M. C. Z. 30761) Tandala, 
Ukinga Mountains. 

Fig. 3. Chamaeleon laterispinis Loveridge, type cf (M. C. Z. 31386) Kigogo, 
Uzungwe Mountains. 

Fig. 4. Chamaeleon incornutus Loveridge, type cf (M. C. Z. 31350) Madehani, 
Ukinga Mountains. 

Fig. 5. Chamaeleon werneri dabagae Loveridge, type 9 (M. C. Z. 31344) 
Dabaga, Uzungwe Mountains. 

All enlarged about 1 7/10 natural size. 



BULL. MUS COMP. ZOOL. 



LovERiDGE. African Herpetology. Pl. 3. 




 -^ I 



}^)1JWt^H 



I 



 Ad^r^iJiK 




THE HEinTYPE CORP. BOSTON 



,v^ 



'J 



(, 



Harvard MCZ Library 




3 2044 066 300 500 



ii^ij. i,. . '-^w.***.© 






^l 



M:. 



r^ 



■</ 



fe-' 



w 



■■^\ 



E^a?wi 


^^^F' * ""^ l^^P^^^^HB^fl\ ^^^^^^Iv ^^^^^1 




B^^y^v > -^ 


3BE>f^ "Jl,tN^ 



V